From ashert at cs.cmu.edu  Fri Feb  4 12:41:53 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Fri, 4 Feb 2022 12:41:53 -0500
Subject: [CMU AI Seminar] Feb 8 at 12pm (Zoom) -- Huan Zhang (CMU) -- How We
 Trust a Black-box: Formal Verification of Deep Neural Networks -- AI Seminar
 sponsored by Morgan Stanley
Message-ID: <CAMVmiv0xP0uxYhfMstMWyELxjpk=BT_obyK3sGEhsFWZbPb3kQ@mail.gmail.com>

Dear all,

Welcome to the CMU AI Seminar for the Spring 2022 semester!

We look forward to seeing you *next Tuesday (2/8)* from *1**2:00-1:00 PM
(U.S. Eastern time)* for the next talk of our *CMU AI seminar*, sponsored by
 Morgan Stanley <https://www.morganstanley.com/about-us/technology/>.

To learn more about the seminar series or see the future schedule, please
visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

On 2/8, *Huan Zhang* (CMU) will be giving a talk titled "*How We Trust a
Black-box: Formal Verification of Deep Neural Networks*" to explain
state-of-the-art neural network verification techniques.

*Title*: How We Trust a Black-box: Formal Verification of Deep Neural
Networks

*Talk Abstract*: Neural networks have become a crucial element in modern
artificial intelligence. However, they are often black-boxes and can behave
unexpectedly and produce surprisingly wrong results. When applying neural
networks to mission-critical systems such as autonomous driving and
aircraft control, it is often desirable to formally verify their
trustworthiness such as safety and robustness. In this talk, I will first
introduce the problem of neural network verification and the challenges
involved to guarantee neural network output given bounded input
perturbations. Then, I will discuss the bound propagation based neural
network verification algorithms such as CROWN and beta-CROWN, which
efficiently propagate linear inequalities through the network in a backward
manner. My talk will highlight state-of-the-art verification techniques
used in our ?,?-CROWN (alpha-beta-CROWN) verifier, a scalable, powerful and
GPU-accelerated neural network verifier that won the 2nd International
Verification of Neural Networks Competition (VNN-COMP?21) with the highest
total score.

*Speaker Bio*: Huan Zhang is a postdoctoral researcher at CMU, supervised
by Prof. Zico Kolter. He received his Ph.D. degree at UCLA in 2020. Huan's
research focuses on the trustworthiness of artificial intelligence,
especially on developing formal verification methods to guarantee the
robustness and safety of machine learning. Huan was awarded an IBM Ph.D.
fellowship and he led the winning team in the 2021 International
Verification of Neural Networks Competition. Huan received the 2021 AdvML
Rising Star Award sponsored by MIT-IBM Watson AI Lab.

*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220204/01d5ac56/attachment.html>

From ashert at cs.cmu.edu  Mon Feb  7 12:12:36 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Mon, 7 Feb 2022 12:12:36 -0500
Subject: [CMU AI Seminar] Feb 8 at 12pm (Zoom) -- Huan Zhang (CMU) -- How
 We Trust a Black-box: Formal Verification of Deep Neural Networks --
 AI Seminar sponsored by Morgan Stanley
In-Reply-To: <CAMVmiv0xP0uxYhfMstMWyELxjpk=BT_obyK3sGEhsFWZbPb3kQ@mail.gmail.com>
References: <CAMVmiv0xP0uxYhfMstMWyELxjpk=BT_obyK3sGEhsFWZbPb3kQ@mail.gmail.com>
Message-ID: <CAMVmiv3OfP=M+DTbHHi8Sjk5y+4cvs8+_3QS1Ez_J_pmALRdGQ@mail.gmail.com>

Hi all,

Just a reminder that the CMU AI Seminar <http://www.cs.cmu.edu/~aiseminar/> is
tomorrow *12pm-1pm*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09.

*Huan Zhang (CMU)* will be talking about his award-winning neural network
verification techniques, as well as neural network verification more
generally.

Thanks,
Asher

On Fri, Feb 4, 2022 at 12:41 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Dear all,
>
> Welcome to the CMU AI Seminar for the Spring 2022 semester!
>
> We look forward to seeing you *next Tuesday (2/8)* from *1**2:00-1:00 PM
> (U.S. Eastern time)* for the next talk of our *CMU AI seminar*, sponsored
> by Morgan Stanley <https://www.morganstanley.com/about-us/technology/>.
>
> To learn more about the seminar series or see the future schedule, please
> visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>
> On 2/8, *Huan Zhang* (CMU) will be giving a talk titled "*How We Trust a
> Black-box: Formal Verification of Deep Neural Networks*" to explain
> state-of-the-art neural network verification techniques.
>
> *Title*: How We Trust a Black-box: Formal Verification of Deep Neural
> Networks
>
> *Talk Abstract*: Neural networks have become a crucial element in modern
> artificial intelligence. However, they are often black-boxes and can behave
> unexpectedly and produce surprisingly wrong results. When applying neural
> networks to mission-critical systems such as autonomous driving and
> aircraft control, it is often desirable to formally verify their
> trustworthiness such as safety and robustness. In this talk, I will first
> introduce the problem of neural network verification and the challenges
> involved to guarantee neural network output given bounded input
> perturbations. Then, I will discuss the bound propagation based neural
> network verification algorithms such as CROWN and beta-CROWN, which
> efficiently propagate linear inequalities through the network in a backward
> manner. My talk will highlight state-of-the-art verification techniques
> used in our ?,?-CROWN (alpha-beta-CROWN) verifier, a scalable, powerful and
> GPU-accelerated neural network verifier that won the 2nd International
> Verification of Neural Networks Competition (VNN-COMP?21) with the highest
> total score.
>
> *Speaker Bio*: Huan Zhang is a postdoctoral researcher at CMU, supervised
> by Prof. Zico Kolter. He received his Ph.D. degree at UCLA in 2020. Huan's
> research focuses on the trustworthiness of artificial intelligence,
> especially on developing formal verification methods to guarantee the
> robustness and safety of machine learning. Huan was awarded an IBM Ph.D.
> fellowship and he led the winning team in the 2021 International
> Verification of Neural Networks Competition. Huan received the 2021 AdvML
> Rising Star Award sponsored by MIT-IBM Watson AI Lab.
>
> *Zoom Link*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>
> Thanks,
> Asher Trockman
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220207/c22b34f8/attachment.html>

From ashert at cs.cmu.edu  Tue Feb  8 11:59:40 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 8 Feb 2022 11:59:40 -0500
Subject: [CMU AI Seminar] Feb 8 at 12pm (Zoom) -- Huan Zhang (CMU) -- How
 We Trust a Black-box: Formal Verification of Deep Neural Networks --
 AI Seminar sponsored by Morgan Stanley
In-Reply-To: <CAMVmiv3OfP=M+DTbHHi8Sjk5y+4cvs8+_3QS1Ez_J_pmALRdGQ@mail.gmail.com>
References: <CAMVmiv0xP0uxYhfMstMWyELxjpk=BT_obyK3sGEhsFWZbPb3kQ@mail.gmail.com>
 <CAMVmiv3OfP=M+DTbHHi8Sjk5y+4cvs8+_3QS1Ez_J_pmALRdGQ@mail.gmail.com>
Message-ID: <CAMVmiv1b3yuPr46OuJz38bEH5Da5bSdYknFShBJET5DiGBqS_w@mail.gmail.com>

Hi all,

The seminar today by Huan Zhang on neural network verification is happening
right now!

In case you are interested:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher

On Mon, Feb 7, 2022 at 12:12 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Hi all,
>
> Just a reminder that the CMU AI Seminar
> <http://www.cs.cmu.edu/~aiseminar/> is tomorrow *12pm-1pm*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09.
>
> *Huan Zhang (CMU)* will be talking about his award-winning neural network
> verification techniques, as well as neural network verification more
> generally.
>
> Thanks,
> Asher
>
> On Fri, Feb 4, 2022 at 12:41 PM Asher Trockman <ashert at cs.cmu.edu> wrote:
>
>> Dear all,
>>
>> Welcome to the CMU AI Seminar for the Spring 2022 semester!
>>
>> We look forward to seeing you *next Tuesday (2/8)* from *1**2:00-1:00 PM
>> (U.S. Eastern time)* for the next talk of our *CMU AI seminar*,
>> sponsored by Morgan Stanley
>> <https://www.morganstanley.com/about-us/technology/>.
>>
>> To learn more about the seminar series or see the future schedule, please
>> visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>>
>> On 2/8, *Huan Zhang* (CMU) will be giving a talk titled "*How We Trust a
>> Black-box: Formal Verification of Deep Neural Networks*" to explain
>> state-of-the-art neural network verification techniques.
>>
>> *Title*: How We Trust a Black-box: Formal Verification of Deep Neural
>> Networks
>>
>> *Talk Abstract*: Neural networks have become a crucial element in modern
>> artificial intelligence. However, they are often black-boxes and can behave
>> unexpectedly and produce surprisingly wrong results. When applying neural
>> networks to mission-critical systems such as autonomous driving and
>> aircraft control, it is often desirable to formally verify their
>> trustworthiness such as safety and robustness. In this talk, I will first
>> introduce the problem of neural network verification and the challenges
>> involved to guarantee neural network output given bounded input
>> perturbations. Then, I will discuss the bound propagation based neural
>> network verification algorithms such as CROWN and beta-CROWN, which
>> efficiently propagate linear inequalities through the network in a backward
>> manner. My talk will highlight state-of-the-art verification techniques
>> used in our ?,?-CROWN (alpha-beta-CROWN) verifier, a scalable, powerful and
>> GPU-accelerated neural network verifier that won the 2nd International
>> Verification of Neural Networks Competition (VNN-COMP?21) with the highest
>> total score.
>>
>> *Speaker Bio*: Huan Zhang is a postdoctoral researcher at CMU,
>> supervised by Prof. Zico Kolter. He received his Ph.D. degree at UCLA in
>> 2020. Huan's research focuses on the trustworthiness of artificial
>> intelligence, especially on developing formal verification methods to
>> guarantee the robustness and safety of machine learning. Huan was
>> awarded an IBM Ph.D. fellowship and he led the winning team in the 2021
>> International Verification of Neural Networks Competition. Huan received
>> the 2021 AdvML Rising Star Award sponsored by MIT-IBM Watson AI Lab.
>>
>> *Zoom Link*:
>> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>>
>> Thanks,
>> Asher Trockman
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220208/b2f8586c/attachment-0001.html>

From ashert at cs.cmu.edu  Fri Feb 11 12:20:09 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Fri, 11 Feb 2022 12:20:09 -0500
Subject: [CMU AI Seminar] Feb 15 at 12pm (Zoom) -- Christian Theobalt (Max
 Planck Institute for Informatics) -- Neural Methods for Reconstruction and
 Rendering of Real World Scenes -- AI Seminar sponsored by Morgan Stanley
Message-ID: <CAMVmiv1N0DcqxgObsXaZBzOGwcXsj8r-UBpW534gumt5iwHAxA@mail.gmail.com>

Dear all,

We look forward to seeing you *next Tuesday (2/15)* from *1**2:00-1:00 PM
(U.S. Eastern time)* for the next talk of our *CMU AI seminar*, sponsored
by Morgan Stanley <https://www.morganstanley.com/about-us/technology/>.

To learn more about the seminar series or see the future schedule, please
visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

On 2/15, Professor *Christian Theobalt *(Max Planck Institute for
Informatics) will be giving a talk titled "*Neural Methods for
Reconstruction and Rendering of Real World Scenes*" to share his research
on new neural rendering approaches.

*Title*: Neural Methods for Reconstruction and Rendering of Real World
Scenes

*Talk Abstract*: In this presentation, I will talk about some of the recent
work we did on new methods for reconstructing computer graphics models of
real world scenes from sparse or even monocular video data. These methods
are based on bringing together neural network-based and explicit
model-based approaches. I will also talk about new neural rendering
approaches that combine explicit model-based and neural network based
concepts for image formation in new ways. They enable new means to
synthesize highly realistic imagery and videos of real work scenes under
user control.

*Speaker Bio*: Christian Theobalt is The Scientific Director of the new
Visual Computing and Artificial Intelligence Department at the
Max-Planck-Institute for Informatics, Saarbr?cken, Germany. He is also a
Professor of Computer Science at Saarland University, Germany. From 2007
until 2009 he was a Visiting Assistant Professor in the Department of
Computer Science at Stanford University. He received his MSc degree in
Artificial Intelligence from the University of Edinburgh, his Diplom (MS)
degree in Computer Science from Saarland University, and his PhD (Dr.-Ing.)
from the Max-Planck-Institute for Informatics.

In his research he looks at algorithmic problems that lie at the
intersection of Computer Graphics, Computer Vision and Machine Learning,
such as: static and dynamic 3D scene reconstruction, neural rendering and
neural scene representations, marker-less motion and performance capture,
virtual humans, virtual and augmented reality, computer animation,
intrinsic video and inverse rendering, computational videography, machine
learning for graphics and vision, new sensors for 3D acquisition, as well
as image- and physically-based rendering. He is also interested in using
reconstruction techniques for human computer interaction. For his work, he
received several awards, including the Otto Hahn Medal of the Max-Planck
Society in 2007, the EUROGRAPHICS Young Researcher Award in 2009, the
German Pattern Recognition Award 2012, and the Karl Heinz Beckurts Award in
2017, and the EUROGRAPHICS Outstanding Technical Contributions Award in
2020. He received two ERC grants, an ERC Starting Grant in 2013 and an ERC
Consolidator Grant in 2017.

*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220211/dd881c09/attachment.html>

From ashert at cs.cmu.edu  Mon Feb 14 13:51:03 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Mon, 14 Feb 2022 13:51:03 -0500
Subject: [CMU AI Seminar] Feb 15 at 12pm (Zoom) -- Christian Theobalt (Max
 Planck Institute for Informatics) -- Neural Methods for
 Reconstruction and
 Rendering of Real World Scenes -- AI Seminar sponsored by Morgan Stanley
In-Reply-To: <CAMVmiv1N0DcqxgObsXaZBzOGwcXsj8r-UBpW534gumt5iwHAxA@mail.gmail.com>
References: <CAMVmiv1N0DcqxgObsXaZBzOGwcXsj8r-UBpW534gumt5iwHAxA@mail.gmail.com>
Message-ID: <CAMVmiv2XaA64MECgV+N-ok5EQB0VF_QS7QxKDJe8qsWCijtjqg@mail.gmail.com>

Hi all,

Just a reminder that the CMU AI Seminar <http://www.cs.cmu.edu/~aiseminar/> is
tomorrow *12pm-1pm*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09.

Professor *Christian Theobalt (Max Planck Institute for Informatics)* will
be sharing his research on new neural rendering approaches.

Thanks,
Asher

On Fri, Feb 11, 2022 at 12:20 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Dear all,
>
> We look forward to seeing you *next Tuesday (2/15)* from *1**2:00-1:00 PM
> (U.S. Eastern time)* for the next talk of our *CMU AI seminar*, sponsored
> by Morgan Stanley <https://www.morganstanley.com/about-us/technology/>.
>
> To learn more about the seminar series or see the future schedule, please
> visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>
> On 2/15, Professor *Christian Theobalt *(Max Planck Institute for
> Informatics) will be giving a talk titled "*Neural Methods for
> Reconstruction and Rendering of Real World Scenes*" to share his research
> on new neural rendering approaches.
>
> *Title*: Neural Methods for Reconstruction and Rendering of Real World
> Scenes
>
> *Talk Abstract*: In this presentation, I will talk about some of the
> recent work we did on new methods for reconstructing computer graphics
> models of real world scenes from sparse or even monocular video data. These
> methods are based on bringing together neural network-based and explicit
> model-based approaches. I will also talk about new neural rendering
> approaches that combine explicit model-based and neural network based
> concepts for image formation in new ways. They enable new means to
> synthesize highly realistic imagery and videos of real work scenes under
> user control.
>
> *Speaker Bio*: Christian Theobalt is The Scientific Director of the new
> Visual Computing and Artificial Intelligence Department at the
> Max-Planck-Institute for Informatics, Saarbr?cken, Germany. He is also a
> Professor of Computer Science at Saarland University, Germany. From 2007
> until 2009 he was a Visiting Assistant Professor in the Department of
> Computer Science at Stanford University. He received his MSc degree in
> Artificial Intelligence from the University of Edinburgh, his Diplom (MS)
> degree in Computer Science from Saarland University, and his PhD (Dr.-Ing.)
> from the Max-Planck-Institute for Informatics.
>
> In his research he looks at algorithmic problems that lie at the
> intersection of Computer Graphics, Computer Vision and Machine Learning,
> such as: static and dynamic 3D scene reconstruction, neural rendering and
> neural scene representations, marker-less motion and performance capture,
> virtual humans, virtual and augmented reality, computer animation,
> intrinsic video and inverse rendering, computational videography, machine
> learning for graphics and vision, new sensors for 3D acquisition, as well
> as image- and physically-based rendering. He is also interested in using
> reconstruction techniques for human computer interaction. For his work, he
> received several awards, including the Otto Hahn Medal of the Max-Planck
> Society in 2007, the EUROGRAPHICS Young Researcher Award in 2009, the
> German Pattern Recognition Award 2012, and the Karl Heinz Beckurts Award in
> 2017, and the EUROGRAPHICS Outstanding Technical Contributions Award in
> 2020. He received two ERC grants, an ERC Starting Grant in 2013 and an ERC
> Consolidator Grant in 2017.
>
> *Zoom Link*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>
> Thanks,
> Asher Trockman
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220214/3adb4df8/attachment.html>

From ashert at cs.cmu.edu  Tue Feb 15 11:59:46 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 15 Feb 2022 11:59:46 -0500
Subject: [CMU AI Seminar] Feb 15 at 12pm (Zoom) -- Christian Theobalt (Max
 Planck Institute for Informatics) -- Neural Methods for
 Reconstruction and
 Rendering of Real World Scenes -- AI Seminar sponsored by Morgan Stanley
In-Reply-To: <CAMVmiv2XaA64MECgV+N-ok5EQB0VF_QS7QxKDJe8qsWCijtjqg@mail.gmail.com>
References: <CAMVmiv1N0DcqxgObsXaZBzOGwcXsj8r-UBpW534gumt5iwHAxA@mail.gmail.com>
 <CAMVmiv2XaA64MECgV+N-ok5EQB0VF_QS7QxKDJe8qsWCijtjqg@mail.gmail.com>
Message-ID: <CAMVmiv1-Rj8Cb1i37+nEseWEaq0tYrwHVcrOVDdPh7r6yJ0Fag@mail.gmail.com>

Hi all,

The seminar today by Christian Theobalt on neural rendering is happening
right now!

In case you are interested:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher

On Mon, Feb 14, 2022 at 1:51 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Hi all,
>
> Just a reminder that the CMU AI Seminar
> <http://www.cs.cmu.edu/~aiseminar/> is tomorrow *12pm-1pm*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09.
>
> Professor *Christian Theobalt (Max Planck Institute for Informatics)* will
> be sharing his research on new neural rendering approaches.
>
> Thanks,
> Asher
>
> On Fri, Feb 11, 2022 at 12:20 PM Asher Trockman <ashert at cs.cmu.edu> wrote:
>
>> Dear all,
>>
>> We look forward to seeing you *next Tuesday (2/15)* from *1**2:00-1:00
>> PM (U.S. Eastern time)* for the next talk of our *CMU AI seminar*,
>> sponsored by Morgan Stanley
>> <https://www.morganstanley.com/about-us/technology/>.
>>
>> To learn more about the seminar series or see the future schedule, please
>> visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>>
>> On 2/15, Professor *Christian Theobalt *(Max Planck Institute for
>> Informatics) will be giving a talk titled "*Neural Methods for
>> Reconstruction and Rendering of Real World Scenes*" to share his
>> research on new neural rendering approaches.
>>
>> *Title*: Neural Methods for Reconstruction and Rendering of Real World
>> Scenes
>>
>> *Talk Abstract*: In this presentation, I will talk about some of the
>> recent work we did on new methods for reconstructing computer graphics
>> models of real world scenes from sparse or even monocular video data. These
>> methods are based on bringing together neural network-based and explicit
>> model-based approaches. I will also talk about new neural rendering
>> approaches that combine explicit model-based and neural network based
>> concepts for image formation in new ways. They enable new means to
>> synthesize highly realistic imagery and videos of real work scenes under
>> user control.
>>
>> *Speaker Bio*: Christian Theobalt is The Scientific Director of the new
>> Visual Computing and Artificial Intelligence Department at the
>> Max-Planck-Institute for Informatics, Saarbr?cken, Germany. He is also a
>> Professor of Computer Science at Saarland University, Germany. From 2007
>> until 2009 he was a Visiting Assistant Professor in the Department of
>> Computer Science at Stanford University. He received his MSc degree in
>> Artificial Intelligence from the University of Edinburgh, his Diplom (MS)
>> degree in Computer Science from Saarland University, and his PhD (Dr.-Ing.)
>> from the Max-Planck-Institute for Informatics.
>>
>> In his research he looks at algorithmic problems that lie at the
>> intersection of Computer Graphics, Computer Vision and Machine Learning,
>> such as: static and dynamic 3D scene reconstruction, neural rendering and
>> neural scene representations, marker-less motion and performance capture,
>> virtual humans, virtual and augmented reality, computer animation,
>> intrinsic video and inverse rendering, computational videography, machine
>> learning for graphics and vision, new sensors for 3D acquisition, as well
>> as image- and physically-based rendering. He is also interested in using
>> reconstruction techniques for human computer interaction. For his work, he
>> received several awards, including the Otto Hahn Medal of the Max-Planck
>> Society in 2007, the EUROGRAPHICS Young Researcher Award in 2009, the
>> German Pattern Recognition Award 2012, and the Karl Heinz Beckurts Award in
>> 2017, and the EUROGRAPHICS Outstanding Technical Contributions Award in
>> 2020. He received two ERC grants, an ERC Starting Grant in 2013 and an ERC
>> Consolidator Grant in 2017.
>>
>> *Zoom Link*:
>> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>>
>> Thanks,
>> Asher Trockman
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220215/68675b0e/attachment.html>

From ashert at cs.cmu.edu  Mon Feb 21 16:40:17 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Mon, 21 Feb 2022 16:40:17 -0500
Subject: [CMU AI Seminar] Feb 22 at 12pm (Zoom) -- Gaurav Manek (CMU) -- Hough
 and Cover: 2D Bin Packing with Classic Computer Vision Techniques --
 AI Seminar sponsored by Morgan Stanley
Message-ID: <CAMVmiv0MN+x5X-OKtBFxQ36DE2Te8EdSSUUVDnimo-nAHYq3JA@mail.gmail.com>

Dear all,

We look forward to seeing you *tomorrow,* *this Tuesday (2/22)* from
*1**2:00-1:00
PM (U.S. Eastern time)* for the next talk of our *CMU AI seminar*,
sponsored by Morgan Stanley
<https://www.morganstanley.com/about-us/technology/>.

To learn more about the seminar series or see the future schedule, please
visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

Tomorrow (2/22), *Gaurav Manek *(CMU) will be giving a talk titled *"**Hough
and Cover: 2D Bin Packing with Classic Computer Vision Techniques**".*

*Title*: Hough and Cover: 2D Bin Packing with Classic Computer Vision
Techniques

*Talk Abstract*: Classic computer vision has, unfortunately, fallen by the
wayside in favor of neural networks. In this talk I introduce a novel
use-case for some such techniques, applying them to 2D bin-packing. The 2D
bin packing problem is a classic NP-hard optimization problem in which a
number of rectangles need to be packed into a minimal number of larger
rectangular bins. I present Hough and Cover, a technique that leverages
transformations found in classic computer vision to greatly reduce the
number of constraints of a novel integer linear programming (ILP)
formulation. This approach is evaluated on classic datasets against a
state-of-the-art ILP approach, where it achieves a speed-up of 5x in many
cases. This problem has many industrial applications such as wood or glass
industries, sheet metal fabrication, and even in scheduling heterogeneous
resources. Solving this optimization problem optimally is important since
it can save money, human resources, and time.
*Presented in Partial Fulfillment of the CSD Speaking Skills Requirement.*

*Speaker Bio*: Gaurav Manek is a PhD student at Carnegie Mellon University,
advised by Zico Kolter. Gaurav's research centers around novel ways to
learn dynamics models and value functions for reinforcement learning and
similar applications, and to understand the subtle mechanisms at work when
these models fail. His interests stretch from theoretical contributions to
industry-facing applications. His personal website is at
https://www.gauravmanek.com/.

*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220221/150f4324/attachment.html>

From ashert at cs.cmu.edu  Tue Feb 22 12:00:55 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 22 Feb 2022 12:00:55 -0500
Subject: [CMU AI Seminar] Feb 22 at 12pm (Zoom) -- Gaurav Manek (CMU) --
 Hough and Cover: 2D Bin Packing with Classic Computer Vision Techniques -- AI
 Seminar sponsored by Morgan Stanley
In-Reply-To: <CAMVmiv0MN+x5X-OKtBFxQ36DE2Te8EdSSUUVDnimo-nAHYq3JA@mail.gmail.com>
References: <CAMVmiv0MN+x5X-OKtBFxQ36DE2Te8EdSSUUVDnimo-nAHYq3JA@mail.gmail.com>
Message-ID: <CAMVmiv0yy0Pz1m0hf41c3z5fhZQqGjEK0F=Mp=Lz8aP7sz31bw@mail.gmail.com>

Hi all,

The seminar today by Gaurav Manek on the use of computer vision techniques
to solve bin packing problems is happening right now!

In case you are interested:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher

On Mon, Feb 21, 2022 at 4:40 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Dear all,
>
> We look forward to seeing you *tomorrow,* *this Tuesday (2/22)* from *1**2:00-1:00
> PM (U.S. Eastern time)* for the next talk of our *CMU AI seminar*,
> sponsored by Morgan Stanley
> <https://www.morganstanley.com/about-us/technology/>.
>
> To learn more about the seminar series or see the future schedule, please
> visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>
> Tomorrow (2/22), *Gaurav Manek *(CMU) will be giving a talk titled *"**Hough
> and Cover: 2D Bin Packing with Classic Computer Vision Techniques**".*
>
> *Title*: Hough and Cover: 2D Bin Packing with Classic Computer Vision
> Techniques
>
> *Talk Abstract*: Classic computer vision has, unfortunately, fallen by
> the wayside in favor of neural networks. In this talk I introduce a novel
> use-case for some such techniques, applying them to 2D bin-packing. The 2D
> bin packing problem is a classic NP-hard optimization problem in which a
> number of rectangles need to be packed into a minimal number of larger
> rectangular bins. I present Hough and Cover, a technique that leverages
> transformations found in classic computer vision to greatly reduce the
> number of constraints of a novel integer linear programming (ILP)
> formulation. This approach is evaluated on classic datasets against a
> state-of-the-art ILP approach, where it achieves a speed-up of 5x in many
> cases. This problem has many industrial applications such as wood or glass
> industries, sheet metal fabrication, and even in scheduling heterogeneous
> resources. Solving this optimization problem optimally is important since
> it can save money, human resources, and time.
> *Presented in Partial Fulfillment of the CSD Speaking Skills Requirement.*
>
> *Speaker Bio*: Gaurav Manek is a PhD student at Carnegie Mellon
> University, advised by Zico Kolter. Gaurav's research centers around
> novel ways to learn dynamics models and value functions for reinforcement
> learning and similar applications, and to understand the subtle mechanisms
> at work when these models fail. His interests stretch from theoretical
> contributions to industry-facing applications. His personal website is at
> https://www.gauravmanek.com/.
>
> *Zoom Link*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>
> Thanks,
> Asher Trockman
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220222/b2d0a81a/attachment.html>

From ashert at cs.cmu.edu  Fri Feb 25 13:41:45 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Fri, 25 Feb 2022 13:41:45 -0500
Subject: [CMU AI Seminar] March 1 at 12pm (Zoom) -- Douwe Kiela (Hugging Face)
 -- Dynabench: Rethinking Benchmarking in AI -- AI Seminar sponsored
 by Morgan Stanley
Message-ID: <CAMVmiv1cvhrp6-5fn2uoS-FY9uDy0eq_bernwCLnVsURsAkG2A@mail.gmail.com>

Dear all,

We look forward to seeing you *next Tuesday (3/1)* from *1**2:00-1:00 PM
(U.S. Eastern time)* for the next talk of our *CMU AI seminar*, sponsored
by Morgan Stanley <https://www.morganstanley.com/about-us/technology/>.

To learn more about the seminar series or see the future schedule, please
visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

On 3/1, *Douwe Kiela *(Hugging Face) will be giving a talk titled
*"**Dynabench:
Rethinking Benchmarking in AI**"* to share his work on addressing problems
with the current benchmarking paradigm in AI.

*Title*: Dynabench: Rethinking Benchmarking in AI

*Talk Abstract*: The current benchmarking paradigm in AI has many issues:
benchmarks saturate quickly, are susceptible to overfitting, contain
exploitable annotator artifacts, have unclear or imperfect evaluation
metrics, and do not necessarily measure what we really care about. I will
talk about our work in trying to rethink the way we do benchmarking in AI,
specifically in natural language processing, focusing mostly on the
Dynabench platform (dynabench.org).

*Speaker Bio*: Douwe Kiela (@douwekiela, https://douwekiela.github.io/) is
the Head of Research at Hugging Face. Before, he was a Research Scientist
at Facebook AI Research. His current research interests lie in developing
better models for (grounded, multi-agent) language understanding and better
tools for evaluation and benchmarking.

*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220225/da011fe4/attachment.html>

From ashert at cs.cmu.edu  Mon Feb 28 16:27:13 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Mon, 28 Feb 2022 16:27:13 -0500
Subject: [CMU AI Seminar] March 1 at 12pm (Zoom) -- Douwe Kiela (Hugging
 Face) -- Dynabench: Rethinking Benchmarking in AI -- AI Seminar sponsored by
 Morgan Stanley
In-Reply-To: <CAMVmiv1cvhrp6-5fn2uoS-FY9uDy0eq_bernwCLnVsURsAkG2A@mail.gmail.com>
References: <CAMVmiv1cvhrp6-5fn2uoS-FY9uDy0eq_bernwCLnVsURsAkG2A@mail.gmail.com>
Message-ID: <CAMVmiv13F46NRTAzEry=35rMxRnTwfd-RXigSPLLh8EX+w0+oQ@mail.gmail.com>

Hi all,

Just a reminder that the CMU AI Seminar <http://www.cs.cmu.edu/~aiseminar/> is
tomorrow *12pm-1pm*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09.

*Douwe Kiela (Hugging Face)* will be sharing his work on addressing
problems with the current benchmarking paradigm in AI.

Thanks,
Asher

On Fri, Feb 25, 2022 at 1:41 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Dear all,
>
> We look forward to seeing you *next Tuesday (3/1)* from *1**2:00-1:00 PM
> (U.S. Eastern time)* for the next talk of our *CMU AI seminar*, sponsored
> by Morgan Stanley <https://www.morganstanley.com/about-us/technology/>.
>
> To learn more about the seminar series or see the future schedule, please
> visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>
> On 3/1, *Douwe Kiela *(Hugging Face) will be giving a talk titled *"**Dynabench:
> Rethinking Benchmarking in AI**"* to share his work on addressing
> problems with the current benchmarking paradigm in AI.
>
> *Title*: Dynabench: Rethinking Benchmarking in AI
>
> *Talk Abstract*: The current benchmarking paradigm in AI has many issues:
> benchmarks saturate quickly, are susceptible to overfitting, contain
> exploitable annotator artifacts, have unclear or imperfect evaluation
> metrics, and do not necessarily measure what we really care about. I will
> talk about our work in trying to rethink the way we do benchmarking in AI,
> specifically in natural language processing, focusing mostly on the
> Dynabench platform (dynabench.org).
>
> *Speaker Bio*: Douwe Kiela (@douwekiela, https://douwekiela.github.io/)
> is the Head of Research at Hugging Face. Before, he was a Research
> Scientist at Facebook AI Research. His current research interests lie in
> developing better models for (grounded, multi-agent) language understanding
> and better tools for evaluation and benchmarking.
>
> *Zoom Link*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>
> Thanks,
> Asher Trockman
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220228/b758cd11/attachment.html>

From nihars at cs.cmu.edu  Mon Feb 28 07:09:21 2022
From: nihars at cs.cmu.edu (Nihar Shah)
Date: Mon, 28 Feb 2022 07:09:21 -0500
Subject: Talk of possible interest: 11am on Friday Mar 4
Message-ID: <CAF7+7kpBEbYov1XPtqYUyhJ8JDrpmO5o=Jr7m7vC1N0UzVugjg@mail.gmail.com>

Dear all,

Wolfgang Kerzendorf <https://wkerzendorf.github.io/#main>, who works in
astrophysics and machine learning, is giving a remote talk in my group
meeting at 11am on Friday March 4 (details of the talk below). I thought
this talk may be of interest to some of you, and if so, you are welcome to
join.
https://cmu.zoom.us/j/99367371591?pwd=SU1jNDZKbU03THJkeEpWVWJtZGV5dz09
(sign in to zoom with a CMU/Andrew account to join)

Title: Distributed peer review enhanced with machine learning

Abstract: Many aspects of modern peer review have not changed from its
inception in the 18th century despite drastic changes in the scientific
community. Specifically, contrarily to the early days of peer review, it
has become a significant challenge to identify experts that can effectively
review the more and more specialized fields of science. The problem is
exacerbated by the ever-rising number of researchers (having grown by 15%
between 2014 and 2018 according to a UNESCO report) also seen through the
staggering increase of publications and proposals (doubling every 14 years
in astronomy) Some say that peer review has not adequately innovated as
technology has advanced and the dissemination of publications has surged,
creating a space for stagnant and biased reviews. We have developed and
deployed a novel form of peer review in which the proposers become
reviewers themselves known as distributed peer review. We enhanced this
process using natural language processing and AI technologies to find an
optimal match between the pool of proposals and the reviewers. In this
talk, I will present this potential solution that was trialed at ESO and is
about to go into full use. I will discuss potential other applications of
AI in the field of peer review. I will close with an outlook of current and
future experiments in peer review.

Best,
Nihar
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220228/d0f6cdb0/attachment.html>

From ashert at cs.cmu.edu  Tue Mar  1 12:01:24 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 1 Mar 2022 12:01:24 -0500
Subject: [CMU AI Seminar] March 1 at 12pm (Zoom) -- Douwe Kiela (Hugging
 Face) -- Dynabench: Rethinking Benchmarking in AI -- AI Seminar sponsored by
 Morgan Stanley
In-Reply-To: <CAMVmiv13F46NRTAzEry=35rMxRnTwfd-RXigSPLLh8EX+w0+oQ@mail.gmail.com>
References: <CAMVmiv1cvhrp6-5fn2uoS-FY9uDy0eq_bernwCLnVsURsAkG2A@mail.gmail.com>
 <CAMVmiv13F46NRTAzEry=35rMxRnTwfd-RXigSPLLh8EX+w0+oQ@mail.gmail.com>
Message-ID: <CAMVmiv1aqf0VGFD71OzZXiJRXmZ8HpCQ3r2-ijMnymLt3wVEHg@mail.gmail.com>

Hi all,

The seminar today by Douwe Kiela on the Dynabench platform is happening
right now!

In case you are interested:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher

On Mon, Feb 28, 2022 at 4:27 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Hi all,
>
> Just a reminder that the CMU AI Seminar
> <http://www.cs.cmu.edu/~aiseminar/> is tomorrow *12pm-1pm*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09.
>
> *Douwe Kiela (Hugging Face)* will be sharing his work on addressing
> problems with the current benchmarking paradigm in AI.
>
> Thanks,
> Asher
>
> On Fri, Feb 25, 2022 at 1:41 PM Asher Trockman <ashert at cs.cmu.edu> wrote:
>
>> Dear all,
>>
>> We look forward to seeing you *next Tuesday (3/1)* from *1**2:00-1:00 PM
>> (U.S. Eastern time)* for the next talk of our *CMU AI seminar*,
>> sponsored by Morgan Stanley
>> <https://www.morganstanley.com/about-us/technology/>.
>>
>> To learn more about the seminar series or see the future schedule, please
>> visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>>
>> On 3/1, *Douwe Kiela *(Hugging Face) will be giving a talk titled *"**Dynabench:
>> Rethinking Benchmarking in AI**"* to share his work on addressing
>> problems with the current benchmarking paradigm in AI.
>>
>> *Title*: Dynabench: Rethinking Benchmarking in AI
>>
>> *Talk Abstract*: The current benchmarking paradigm in AI has many
>> issues: benchmarks saturate quickly, are susceptible to overfitting,
>> contain exploitable annotator artifacts, have unclear or imperfect
>> evaluation metrics, and do not necessarily measure what we really care
>> about. I will talk about our work in trying to rethink the way we do
>> benchmarking in AI, specifically in natural language processing, focusing
>> mostly on the Dynabench platform (dynabench.org).
>>
>> *Speaker Bio*: Douwe Kiela (@douwekiela, https://douwekiela.github.io/)
>> is the Head of Research at Hugging Face. Before, he was a Research
>> Scientist at Facebook AI Research. His current research interests lie in
>> developing better models for (grounded, multi-agent) language understanding
>> and better tools for evaluation and benchmarking.
>>
>> *Zoom Link*:
>> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>>
>> Thanks,
>> Asher Trockman
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220301/e8534eb3/attachment.html>

From ashert at cs.cmu.edu  Mon Mar 14 12:36:24 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Mon, 14 Mar 2022 12:36:24 -0400
Subject: [CMU AI Seminar] March 15 at 12pm (Zoom) -- Kevin Ellis (Cornell) --
 What Program Synthesis Can Learn From How People Write Code -- AI
 Seminar sponsored by Morgan Stanley
Message-ID: <CAMVmiv3LmSL-Pcp2v_Oar1XWOJoCF-4fcoO=SC0v997gg8iSXg@mail.gmail.com>

Dear all,

We look forward to seeing you *tomorrow, this Tuesday (3/15)* from
*1**2:00-1:00
PM (U.S. Eastern time)* for the next talk of our *CMU AI seminar*,
sponsored by Morgan Stanley
<https://www.morganstanley.com/about-us/technology/>.

To learn more about the seminar series or see the future schedule, please
visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

*Tomorrow* (3/15), *Kevin Ellis *(Cornell) will be giving a talk
titled *"**What
Program Synthesis Can Learn From How People Write Code**" *to share his
work on advancing program synthesis using insights from how humans build
software, such as writing libraries and using REPLs.

*Title*: What Program Synthesis Can Learn From How People Write Code

*Talk Abstract*: How can we best make systems which learn to write computer
programs? Here I explore the idea that we should take insight from the
techniques and tools that human coders use when building software, but that
we should combine those insights with machine learning methods. I focus on
two basic coding techniques: writing libraries, and using interpreters
("REPLs"). For libraries, I present a system called DreamCoder, which grows
a library of reusable subroutines as it solves a range of programming
problems. DreamCoder's architecture builds on the structure of wake-sleep
neural network training algorithms, and combines both symbolic and neural
learning. For interpreters, I present a system which learns to interact
with a REPL while it writes code, showing that this can help mitigate the
combinatorial search difficulties of program synthesis. At the end of the
talk, I will present preliminary results on modeling another aspect of
coding: creating challenging and interesting programming problems.

*Speaker Bio*: Kevin Ellis <https://www.cs.cornell.edu/~ellisk/> works in
artificial intelligence and programming languages. He is an assistant
professor of computer science at Cornell University, and was previously a
research scientist at Common Sense Machines. He did his PhD at MIT in
cognitive science.

*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220314/1ac72af7/attachment.html>

From ashert at cs.cmu.edu  Tue Mar 15 12:00:04 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 15 Mar 2022 12:00:04 -0400
Subject: [CMU AI Seminar] March 15 at 12pm (Zoom) -- Kevin Ellis (Cornell)
 -- What Program Synthesis Can Learn From How People Write Code -- AI
 Seminar sponsored by Morgan Stanley
In-Reply-To: <CAMVmiv3LmSL-Pcp2v_Oar1XWOJoCF-4fcoO=SC0v997gg8iSXg@mail.gmail.com>
References: <CAMVmiv3LmSL-Pcp2v_Oar1XWOJoCF-4fcoO=SC0v997gg8iSXg@mail.gmail.com>
Message-ID: <CAMVmiv34-b=tzE189u-yqVL_T6DwuEva7sgL+V9p0wuO0vuvpw@mail.gmail.com>

Hi all,

The seminar today by Kevin Ellis on enhancing program synthesis using
human-inspired techniques is happening right now!

In case you are interested:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher

On Mon, Mar 14, 2022 at 12:36 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Dear all,
>
> We look forward to seeing you *tomorrow, this Tuesday (3/15)* from *1**2:00-1:00
> PM (U.S. Eastern time)* for the next talk of our *CMU AI seminar*,
> sponsored by Morgan Stanley
> <https://www.morganstanley.com/about-us/technology/>.
>
> To learn more about the seminar series or see the future schedule, please
> visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>
> *Tomorrow* (3/15), *Kevin Ellis *(Cornell) will be giving a talk titled
> *"**What Program Synthesis Can Learn From How People Write Code**" *to share
> his work on advancing program synthesis using insights from how humans
> build software, such as writing libraries and using REPLs.
>
> *Title*: What Program Synthesis Can Learn From How People Write Code
>
> *Talk Abstract*: How can we best make systems which learn to write
> computer programs? Here I explore the idea that we should take insight from
> the techniques and tools that human coders use when building software, but
> that we should combine those insights with machine learning methods. I
> focus on two basic coding techniques: writing libraries, and using
> interpreters ("REPLs"). For libraries, I present a system called
> DreamCoder, which grows a library of reusable subroutines as it solves a
> range of programming problems. DreamCoder's architecture builds on the
> structure of wake-sleep neural network training algorithms, and combines
> both symbolic and neural learning. For interpreters, I present a system
> which learns to interact with a REPL while it writes code, showing that
> this can help mitigate the combinatorial search difficulties of program
> synthesis. At the end of the talk, I will present preliminary results on
> modeling another aspect of coding: creating challenging and interesting
> programming problems.
>
> *Speaker Bio*: Kevin Ellis <https://www.cs.cornell.edu/~ellisk/> works in
> artificial intelligence and programming languages. He is an assistant
> professor of computer science at Cornell University, and was previously a
> research scientist at Common Sense Machines. He did his PhD at MIT in
> cognitive science.
>
> *Zoom Link*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>
> Thanks,
> Asher Trockman
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220315/22833506/attachment.html>

From ashert at cs.cmu.edu  Fri Mar 18 17:26:46 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Fri, 18 Mar 2022 17:26:46 -0400
Subject: [CMU AI Seminar] March 22 at 12pm (Zoom) -- Chirag Gupta (CMU) --
 Provably calibrating ML classifiers without distributional assumptions -- AI
 Seminar sponsored by Morgan Stanley
Message-ID: <CAMVmiv3GKpe-9M_mPPXe_fk3s2Hzj3TvimN70-c+gLGUb-wdjw@mail.gmail.com>

Dear all,

We look forward to seeing you *next Tuesday (3/22)* from *1**2:00-1:00 PM
(U.S. Eastern time)* for the next talk of our *CMU AI seminar*, sponsored
by Morgan Stanley <https://www.morganstanley.com/about-us/technology/>.

To learn more about the seminar series or see the future schedule, please
visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

On 3/22, *Chirag Gupta *(CMU) will be giving a talk titled
*"**Provably calibrating
ML classifiers without distributional assumptions**"* to share his work on
new notions of calibration in the multiclass setting.

*Title*: Provably calibrating ML classifiers without distributional
assumptions

*Talk Abstract*: Most ML classifiers provide probability scores for the
different classes. What do these scores mean? Probabilistic classifiers are
said to be calibrated if the observed frequencies of labels match the
claimed/reported probabilities. While calibration in the binary
classification setting has been studied since the mid-1900s, there is less
clarity on the right notion of calibration for multiclass classification.
In this talk, I will present recent work where we investigate the
relationship between commonly considered notions of multiclass calibration
and the calibration algorithms used to achieve these notions. We will
discuss our proposed notion of top-label calibration, and the general
framework of multiclass-to-binary (M2B) calibration. We show that any M2B
notion of calibration can be provably achieved, no matter how the data is
distributed. I will present these calibration guarantees as well as
experimental results on calibrating deep learning models. Our proposed
algorithms beat existing algorithms in most situations. Code for this work
is available at https://github.com/aigen/df-posthoc-calibration.

*Speaker Bio*: Chirag Gupta is a fourth-year PhD student in the Machine
Learning Department at CMU, advised by Aaditya Ramdas. He works on
principled methods for uncertainty quantification in classification and
regression problems.

*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220318/33e84a49/attachment.html>

From ashert at cs.cmu.edu  Tue Mar 22 12:02:46 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 22 Mar 2022 12:02:46 -0400
Subject: [CMU AI Seminar] March 22 at 12pm (Zoom) -- Chirag Gupta (CMU) --
 Provably calibrating ML classifiers without distributional
 assumptions -- AI Seminar sponsored by Morgan Stanley
In-Reply-To: <CAMVmiv3GKpe-9M_mPPXe_fk3s2Hzj3TvimN70-c+gLGUb-wdjw@mail.gmail.com>
References: <CAMVmiv3GKpe-9M_mPPXe_fk3s2Hzj3TvimN70-c+gLGUb-wdjw@mail.gmail.com>
Message-ID: <CAMVmiv3=QtqEEHKnnbNBnE01Twzh0R6baFr-Vr3hYr0Hz747fw@mail.gmail.com>

Hi all,

The seminar today by Chirag Gupta is happening right now!

In case you are interested:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher

On Fri, Mar 18, 2022 at 5:26 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Dear all,
>
> We look forward to seeing you *next Tuesday (3/22)* from *1**2:00-1:00 PM
> (U.S. Eastern time)* for the next talk of our *CMU AI seminar*, sponsored
> by Morgan Stanley <https://www.morganstanley.com/about-us/technology/>.
>
> To learn more about the seminar series or see the future schedule, please
> visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>
> On 3/22, *Chirag Gupta *(CMU) will be giving a talk titled *"**Provably calibrating
> ML classifiers without distributional assumptions**"* to share his work
> on new notions of calibration in the multiclass setting.
>
> *Title*: Provably calibrating ML classifiers without distributional
> assumptions
>
> *Talk Abstract*: Most ML classifiers provide probability scores for the
> different classes. What do these scores mean? Probabilistic classifiers are
> said to be calibrated if the observed frequencies of labels match the
> claimed/reported probabilities. While calibration in the binary
> classification setting has been studied since the mid-1900s, there is less
> clarity on the right notion of calibration for multiclass classification.
> In this talk, I will present recent work where we investigate the
> relationship between commonly considered notions of multiclass calibration
> and the calibration algorithms used to achieve these notions. We will
> discuss our proposed notion of top-label calibration, and the general
> framework of multiclass-to-binary (M2B) calibration. We show that any M2B
> notion of calibration can be provably achieved, no matter how the data is
> distributed. I will present these calibration guarantees as well as
> experimental results on calibrating deep learning models. Our proposed
> algorithms beat existing algorithms in most situations. Code for this work
> is available at https://github.com/aigen/df-posthoc-calibration.
>
> *Speaker Bio*: Chirag Gupta is a fourth-year PhD student in the Machine
> Learning Department at CMU, advised by Aaditya Ramdas. He works on
> principled methods for uncertainty quantification in classification and
> regression problems.
>
> *Zoom Link*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>
> Thanks,
> Asher Trockman
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220322/f38974ca/attachment.html>

From ashert at cs.cmu.edu  Fri Apr  1 16:57:50 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Fri, 1 Apr 2022 16:57:50 -0400
Subject: [CMU AI Seminar] April 5 at 12pm (Zoom) -- Uri Alon (CMU) --
 Neuro-Symbolic Language Modeling with Automaton-augmented Retrieval -- AI
 Seminar sponsored by Morgan Stanley
Message-ID: <CAMVmiv1kT=qRJ8a2QVNJOCs6ykoYBAWNzrx4yTz9Z9t1Ffk6Ug@mail.gmail.com>

Dear all,

We look forward to seeing you *next Tuesday (4/5)* from *1**2:00-1:00 PM
(U.S. Eastern time)* for the next talk of our *CMU AI seminar*, sponsored
by Morgan Stanley <https://www.morganstanley.com/about-us/technology/>.

To learn more about the seminar series or see the future schedule, please
visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

On 4/5, *Uri Alon *(CMU) will be giving a talk titled *"Neuro-Symbolic
Language Modeling with Automaton-augmented Retrieval"* to share his work on
improving retrieval-based language modeling.

*Title*: Neuro-Symbolic Language Modeling with Automaton-augmented Retrieval

*Talk Abstract*: Retrieval-based language models (R-LM) model the
probability of natural language text by combining a standard language model
(LM) with examples retrieved from an external datastore at test time. While
effective, a major bottleneck of using these models in practice is the
computationally costly datastore search, which can be performed as
frequently as every time step. In this talk, I will present RetoMaton -
retrieval automaton - which approximates the datastore search, based on (1)
clustering of entries into "states", and (2) state transitions from
previous entries. This effectively results in a weighted finite automaton
built on top of the datastore, instead of representing the datastore as a
flat list. The creation of the automaton is unsupervised, and a RetoMaton
can be constructed from any text collection: either the original training
corpus or from another domain. Traversing this automaton at inference time,
in parallel to the LM inference, reduces its perplexity, or alternatively
saves up to 83% of the nearest neighbor searches over kNN-LM (Khandelwal et
al., 2020), without hurting perplexity.

*Speaker Bio*: Uri Alon is a Postdoctoral Researcher at LTI, working with
Prof. Graham Neubig on NLP and learning from source code. Previously, he
obtained his PhD at the Technion (Israel), where he worked on modeling
programming languages and graphs. Currently, he is also interested in the
synergy of neural models with symbolic components such as retrieval,
programs, and automata. His personal website is at https://urialon.ml. Feel
free to reach out with any questions or comments about the talk.

*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220401/3bd0e8c9/attachment.html>

From ashert at cs.cmu.edu  Tue Apr  5 12:01:48 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 5 Apr 2022 12:01:48 -0400
Subject: [CMU AI Seminar] April 5 at 12pm (Zoom) -- Uri Alon (CMU) --
 Neuro-Symbolic Language Modeling with Automaton-augmented Retrieval -- AI
 Seminar sponsored by Morgan Stanley
In-Reply-To: <CAMVmiv1kT=qRJ8a2QVNJOCs6ykoYBAWNzrx4yTz9Z9t1Ffk6Ug@mail.gmail.com>
References: <CAMVmiv1kT=qRJ8a2QVNJOCs6ykoYBAWNzrx4yTz9Z9t1Ffk6Ug@mail.gmail.com>
Message-ID: <CAMVmiv0ws46hXmFjDzR7k=bB_vVHbmQhgTXM-p-SGHsbpA2jYQ@mail.gmail.com>

Hi all,

The seminar today by Uri Alon on neuro-symbolic language modeling is
happening right now!

In case you are interested:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher

On Fri, Apr 1, 2022 at 4:57 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Dear all,
>
> We look forward to seeing you *next Tuesday (4/5)* from *1**2:00-1:00 PM
> (U.S. Eastern time)* for the next talk of our *CMU AI seminar*, sponsored
> by Morgan Stanley <https://www.morganstanley.com/about-us/technology/>.
>
> To learn more about the seminar series or see the future schedule, please
> visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>
> On 4/5, *Uri Alon *(CMU) will be giving a talk titled *"Neuro-Symbolic
> Language Modeling with Automaton-augmented Retrieval"* to share his work
> on improving retrieval-based language modeling.
>
> *Title*: Neuro-Symbolic Language Modeling with Automaton-augmented
> Retrieval
>
> *Talk Abstract*: Retrieval-based language models (R-LM) model the
> probability of natural language text by combining a standard language model
> (LM) with examples retrieved from an external datastore at test time. While
> effective, a major bottleneck of using these models in practice is the
> computationally costly datastore search, which can be performed as
> frequently as every time step. In this talk, I will present RetoMaton -
> retrieval automaton - which approximates the datastore search, based on (1)
> clustering of entries into "states", and (2) state transitions from
> previous entries. This effectively results in a weighted finite automaton
> built on top of the datastore, instead of representing the datastore as a
> flat list. The creation of the automaton is unsupervised, and a RetoMaton
> can be constructed from any text collection: either the original training
> corpus or from another domain. Traversing this automaton at inference time,
> in parallel to the LM inference, reduces its perplexity, or alternatively
> saves up to 83% of the nearest neighbor searches over kNN-LM (Khandelwal et
> al., 2020), without hurting perplexity.
>
> *Speaker Bio*: Uri Alon is a Postdoctoral Researcher at LTI, working with
> Prof. Graham Neubig on NLP and learning from source code. Previously, he
> obtained his PhD at the Technion (Israel), where he worked on modeling
> programming languages and graphs. Currently, he is also interested in the
> synergy of neural models with symbolic components such as retrieval,
> programs, and automata. His personal website is at https://urialon.ml.
> Feel free to reach out with any questions or comments about the talk.
>
> *Zoom Link*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>
> Thanks,
> Asher Trockman
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220405/b1ce77b4/attachment.html>

From ashert at cs.cmu.edu  Fri Apr  8 13:41:03 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Fri, 8 Apr 2022 13:41:03 -0400
Subject: [CMU AI Seminar] April 12 at 12pm (Zoom) -- Tom Goldstein (University
 of Maryland) -- End-to-end algorithm synthesis with "thinking"
 networks -- AI Seminar sponsored by Morgan Stanley
Message-ID: <CAMVmiv1o88fQsKYAiNqSfxQfeyjr7QFRkga_GM7D7Kzvdj1kNg@mail.gmail.com>

Dear all,

We look forward to seeing you *next Tuesday (4/12)* from *1**2:00-1:00 PM
(U.S. Eastern time)* for the next talk of our *CMU AI seminar*, sponsored
by Morgan Stanley <https://www.morganstanley.com/about-us/technology/>.

To learn more about the seminar series or see the future schedule, please
visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

On 4/12, *Tom Goldstein *(University of Maryland) will be giving a talk
titled *"**End-to-end algorithm synthesis with 'thinking' networks**"*
to survey
adversarial machine learning and to explain his recent work on "thinking
systems".

*Title*: End-to-end algorithm synthesis with "thinking" networks

*Talk Abstract*: This talk will have two parts.  In the first half of the
talk, I'll survey the basics of adversarial machine learning, and discuss
whether adversarial attacks and dataset poisoning can scale up to work on
industrial systems.  I'll also present applications where adversarial
methods provide benefits for domain shift robustness, dataset privacy, and
data augmentation.  In the second half of the talk, I'll present my recent
work on "thinking systems."  These systems use recurrent networks to
emulate a human-like thinking process, in which problems are represented in
memory and then iteratively manipulated and simplified over time until a
solution to a problem is found.  When these models are trained only on
"easy" problem instances, they can then solve "hard" problem instances
without having ever seen one, provided the model is allowed the "think" for
longer at test time.

*Speaker Bio*: Tom Goldstein is the Perotto Associate Professor of Computer
Science at the University of Maryland.  His research lies at the
intersection of machine learning and optimization, and targets applications
in computer vision and signal processing. Before joining the faculty at
Maryland, Tom completed his PhD in Mathematics at UCLA, and was a research
scientist at Rice University and Stanford University. Professor Goldstein
has been the recipient of several awards, including SIAM?s DiPrima Prize, a
DARPA Young Faculty Award, a JP Morgan Faculty award, and a Sloan
Fellowship.

*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220408/b00575a0/attachment.html>

From ashert at cs.cmu.edu  Tue Apr 12 11:28:25 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 12 Apr 2022 11:28:25 -0400
Subject: [CMU AI Seminar] April 12 at 12pm (Zoom) -- Tom Goldstein
 (University of Maryland) -- End-to-end algorithm synthesis with "thinking"
 networks -- AI Seminar sponsored by Morgan Stanley
In-Reply-To: <CAMVmiv1o88fQsKYAiNqSfxQfeyjr7QFRkga_GM7D7Kzvdj1kNg@mail.gmail.com>
References: <CAMVmiv1o88fQsKYAiNqSfxQfeyjr7QFRkga_GM7D7Kzvdj1kNg@mail.gmail.com>
Message-ID: <CAMVmiv1T+n=6uFCCJ+2v4vcg3dzhG_UJUG901GvaOP+JYdJBfA@mail.gmail.com>

Hi all,

Just a reminder that the talk today by Tom Goldstein on "End-to-end
algorithm synthesis with 'thinking' networks" is happening at 12pm!

Zoom Link:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Best,
Asher

On Fri, Apr 8, 2022 at 1:41 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Dear all,
>
> We look forward to seeing you *next Tuesday (4/12)* from *1**2:00-1:00 PM
> (U.S. Eastern time)* for the next talk of our *CMU AI seminar*, sponsored
> by Morgan Stanley <https://www.morganstanley.com/about-us/technology/>.
>
> To learn more about the seminar series or see the future schedule, please
> visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>
> On 4/12, *Tom Goldstein *(University of Maryland) will be giving a talk
> titled *"**End-to-end algorithm synthesis with 'thinking' networks**"* to survey
> adversarial machine learning and to explain his recent work on "thinking
> systems".
>
> *Title*: End-to-end algorithm synthesis with "thinking" networks
>
> *Talk Abstract*: This talk will have two parts.  In the first half of the
> talk, I'll survey the basics of adversarial machine learning, and discuss
> whether adversarial attacks and dataset poisoning can scale up to work on
> industrial systems.  I'll also present applications where adversarial
> methods provide benefits for domain shift robustness, dataset privacy, and
> data augmentation.  In the second half of the talk, I'll present my recent
> work on "thinking systems."  These systems use recurrent networks to
> emulate a human-like thinking process, in which problems are represented in
> memory and then iteratively manipulated and simplified over time until a
> solution to a problem is found.  When these models are trained only on
> "easy" problem instances, they can then solve "hard" problem instances
> without having ever seen one, provided the model is allowed the "think" for
> longer at test time.
>
> *Speaker Bio*: Tom Goldstein is the Perotto Associate Professor of
> Computer Science at the University of Maryland.  His research lies at the
> intersection of machine learning and optimization, and targets applications
> in computer vision and signal processing. Before joining the faculty at
> Maryland, Tom completed his PhD in Mathematics at UCLA, and was a research
> scientist at Rice University and Stanford University. Professor Goldstein
> has been the recipient of several awards, including SIAM?s DiPrima Prize, a
> DARPA Young Faculty Award, a JP Morgan Faculty award, and a Sloan
> Fellowship.
>
> *Zoom Link*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>
> Thanks,
> Asher Trockman
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220412/bd0228fa/attachment.html>

From ashert at cs.cmu.edu  Mon May  9 11:31:02 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Mon, 9 May 2022 10:31:02 -0500
Subject: [CMU AI Seminar] May 10 at 12pm (Zoom) -- Albert Gu (Stanford) --
 Efficiently Modeling Long Sequences with Structured State Spaces -- AI
 Seminar sponsored by Morgan Stanley
Message-ID: <CAMVmiv0M9EyRrF6rPFx4nWyEyvXU13bhxsrDRcY7FpVW8u07+Q@mail.gmail.com>

Dear all,

We look forward to seeing you *tomorrow, this Tuesday (5/10)* from
*1**2:00-1:00
PM (U.S. Eastern time)* for the next talk of our *CMU AI seminar*,
sponsored by Morgan Stanley
<https://www.morganstanley.com/about-us/technology/>.

To learn more about the seminar series or see the future schedule, please
visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

*Tomorrow* (5/10), *Albert Gu *(Stanford) will be giving a talk titled
*"**Efficiently
Modeling Long Sequences with Structured State Spaces**" *to share his work
proposing the S4 model, which handles long-range dependencies
mathematically and empirically, and can be computed very efficiently.

*Title*: Efficiently Modeling Long Sequences with Structured State Spaces

*Talk Abstract*: A central goal of sequence modeling is designing a single
principled model that can address sequence data across a range of
modalities and tasks, particularly on long-range dependencies.  Although
conventional models including RNNs, CNNs, and Transformers have specialized
variants for capturing long dependencies, they still struggle to scale to
very long sequences of 10000 or more steps.  This talk introduces the
Structured State Space sequence model (S4), a simple new model based on the
fundamental state space representation $x'(t) = Ax(t) + Bu(t), y(t) = Cx(t)
+ Du(t)$. S4 combines elegant properties of state space models with the
recent HiPPO theory of continuous-time memorization, resulting in a class
of structured models that handles long-range dependencies mathematically
and can be computed very efficiently. S4 achieves strong empirical results
across a diverse range of established benchmarks, particularly for
continuous signal data such as images, audio, and time series.

*Speaker Bio*: Albert Gu is a final year Ph.D. candidate in the Department
of Computer Science at Stanford University, advised by Christopher R?. His
research broadly studies structured representations for advancing the
capabilities of machine learning and deep learning models, with focuses on
structured linear algebra, non-Euclidean representations, and theory of
sequence models. Previously, he completed a B.S. in Mathematics and
Computer Science at Carnegie Mellon University.

*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220509/ac3a06ff/attachment.html>

From ashert at cs.cmu.edu  Tue May 10 11:50:29 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 10 May 2022 11:50:29 -0400
Subject: [CMU AI Seminar] May 10 at 12pm (Zoom) -- Albert Gu (Stanford) --
 Efficiently Modeling Long Sequences with Structured State Spaces -- AI
 Seminar sponsored by Morgan Stanley
In-Reply-To: <CAMVmiv0M9EyRrF6rPFx4nWyEyvXU13bhxsrDRcY7FpVW8u07+Q@mail.gmail.com>
References: <CAMVmiv0M9EyRrF6rPFx4nWyEyvXU13bhxsrDRcY7FpVW8u07+Q@mail.gmail.com>
Message-ID: <CAMVmiv1jOrrA7GRPbW68tUJvSp9Wn4Ec_tRFfBE_OoHGwqVS+A@mail.gmail.com>

Hi all,

Just a reminder that Albert will be giving his talk on "Efficiently
Modeling Long Sequences with Structured State Spaces" in 10 minutes. Zoom:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher


On Mon, May 9, 2022 at 11:31 AM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Dear all,
>
> We look forward to seeing you *tomorrow, this Tuesday (5/10)* from *1**2:00-1:00
> PM (U.S. Eastern time)* for the next talk of our *CMU AI seminar*,
> sponsored by Morgan Stanley
> <https://www.morganstanley.com/about-us/technology/>.
>
> To learn more about the seminar series or see the future schedule, please
> visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>
> *Tomorrow* (5/10), *Albert Gu *(Stanford) will be giving a talk titled *"**Efficiently
> Modeling Long Sequences with Structured State Spaces**" *to share his
> work proposing the S4 model, which handles long-range dependencies
> mathematically and empirically, and can be computed very efficiently.
>
> *Title*: Efficiently Modeling Long Sequences with Structured State Spaces
>
> *Talk Abstract*: A central goal of sequence modeling is designing a
> single principled model that can address sequence data across a range of
> modalities and tasks, particularly on long-range dependencies.  Although
> conventional models including RNNs, CNNs, and Transformers have specialized
> variants for capturing long dependencies, they still struggle to scale to
> very long sequences of 10000 or more steps.  This talk introduces the
> Structured State Space sequence model (S4), a simple new model based on the
> fundamental state space representation $x'(t) = Ax(t) + Bu(t), y(t) = Cx(t)
> + Du(t)$. S4 combines elegant properties of state space models with the
> recent HiPPO theory of continuous-time memorization, resulting in a class
> of structured models that handles long-range dependencies mathematically
> and can be computed very efficiently. S4 achieves strong empirical results
> across a diverse range of established benchmarks, particularly for
> continuous signal data such as images, audio, and time series.
>
> *Speaker Bio*: Albert Gu is a final year Ph.D. candidate in the
> Department of Computer Science at Stanford University, advised by
> Christopher R?. His research broadly studies structured representations for
> advancing the capabilities of machine learning and deep learning models,
> with focuses on structured linear algebra, non-Euclidean representations,
> and theory of sequence models. Previously, he completed a B.S. in
> Mathematics and Computer Science at Carnegie Mellon University.
>
> *Zoom Link*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>
> Thanks,
> Asher Trockman
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220510/754aa6ef/attachment.html>

From ashert at cs.cmu.edu  Tue May 10 13:41:07 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 10 May 2022 13:41:07 -0400
Subject: [CMU AI Seminar] Special! May 13 at 1:45pm (Hybrid) -- Eric Xing
 (CMU) -- From Learning, to Meta-Learning, to "Lego-Learning" -- theory,
 system, and engineering -- AI Seminar sponsored by Morgan Stanley
Message-ID: <CAMVmiv3AyZMUXGkzCRn1DiN1ZDffuvQuGzMAyGxD=UiGyu=8jQ@mail.gmail.com>

Dear all,

We invite you to a special installment of our CMU AI Seminar Series* this
Friday (5/13)* in *GHC 8102 *and on Zoom from *1:45 - 2:45** PM (U.S.
Eastern time)*, sponsored by Morgan Stanley
<https://www.morganstanley.com/about-us/technology/>.

To learn more about the seminar series or see the future schedule, please
visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

On this Friday (5/13), *Eric Xing *(CMU & MBZUAI) will be giving a talk
titled *"From Learning, to Meta-Learning, to "Lego-Learning" -- theory,
system, and engineering"* to share his work towards building machine
learning pipelines and systems that meet highly-demanding industrial
standards.

*Title*: From Learning, to Meta-Learning, to "Lego-Learning" -- theory,
system, and engineering

*Talk Abstract*: Software systems for complex tasks - such as controlling
manufacturing processes in real-time; or writing radiological case reports
within a clinical workflow ? are becoming increasingly sophisticated and
consist of a large number of data, model, algorithm, and system elements
and modules. Traditional benchmark/leaderboard-driven bespoke approaches in
the Machine Learning community are not suited to meet the highly demanding
industrial standards beyond algorithmic performance, such as
cost-effectiveness, safety, scalability, and automatability, typically
expected in production systems. In this talk, I discuss some technical
issues toward addressing these challenges: 1) a theoretical framework for
panoramic learning with all experiences; 2) optimization methods to best
the effort for learning under such a principled framework; 3) compositional
strategies for building production-grade ML programs from standard parts. I
will present our recent work toward developing a standard model for
Learning that unifies different special-purpose machine learning paradigms
and algorithms, then a Bayesian blackbox optimization approach to Meta
Learning in the space of hyperparameters, model architectures, and system
configurations, and finally principles and designs of standardized software
Legos that facilitate cost-effective building, training, and tuning of
practical ML pipelines and systems.

*Speaker Bio*: Eric P. Xing is the President of the Mohamed bin Zayed
University of Artificial Intelligence, a Professor of Computer Science at
Carnegie Mellon University, and the Founder and Chairman of Petuum Inc., a
2018 World Economic Forum Technology Pioneer company that builds
standardized artificial intelligence development platform and operating
system for broad and general industrial AI applications. He completed his
PhD in Computer Science at UC Berkeley. His main research interests are the
development of machine learning and statistical methodology; and
composable, automatic, and scalable computational systems, for solving
problems involving automated learning, reasoning, and decision-making in
artificial, biological, and social systems. Prof. Xing currently serves or
has served the following roles: associate editor of the Journal of the
American Statistical Association (JASA), Annals of Applied Statistics
(AOAS), and IEEE Journal of Pattern Analysis and Machine Intelligence
(PAMI); action editor of the Machine Learning Journal (MLJ) and Journal of
Machine Learning Research (JMLR); he is a board member of the International
Machine Learning Society.

*In Person: *GHC 8102
*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220510/eb2e2653/attachment.html>

From ashert at cs.cmu.edu  Fri May 13 12:40:33 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Fri, 13 May 2022 12:40:33 -0400
Subject: [CMU AI Seminar] Special! May 13 at 1:45pm (Hybrid) -- Eric Xing
 (CMU) -- From Learning, to Meta-Learning, to "Lego-Learning" -- theory,
 system, and engineering -- AI Seminar sponsored by Morgan Stanley
In-Reply-To: <CAMVmiv3AyZMUXGkzCRn1DiN1ZDffuvQuGzMAyGxD=UiGyu=8jQ@mail.gmail.com>
References: <CAMVmiv3AyZMUXGkzCRn1DiN1ZDffuvQuGzMAyGxD=UiGyu=8jQ@mail.gmail.com>
Message-ID: <CAMVmiv3wUis5FD+sUoyfAofy-J_K6UZYVRw8wC45ALUUES9UNQ@mail.gmail.com>

Hi all,

Just a reminder that Eric Xing will be giving a talk today at 1:45pm in GHC
8102. You can also join on Zoom:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher

On Tue, May 10, 2022 at 1:41 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Dear all,
>
> We invite you to a special installment of our CMU AI Seminar Series* this
> Friday (5/13)* in *GHC 8102 *and on Zoom from *1:45 - 2:45** PM (U.S.
> Eastern time)*, sponsored by Morgan Stanley
> <https://www.morganstanley.com/about-us/technology/>.
>
> To learn more about the seminar series or see the future schedule, please
> visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>
> On this Friday (5/13), *Eric Xing *(CMU & MBZUAI) will be giving a talk
> titled *"From Learning, to Meta-Learning, to "Lego-Learning" -- theory,
> system, and engineering"* to share his work towards building machine
> learning pipelines and systems that meet highly-demanding industrial
> standards.
>
> *Title*: From Learning, to Meta-Learning, to "Lego-Learning" -- theory,
> system, and engineering
>
> *Talk Abstract*: Software systems for complex tasks - such as controlling
> manufacturing processes in real-time; or writing radiological case reports
> within a clinical workflow ? are becoming increasingly sophisticated and
> consist of a large number of data, model, algorithm, and system elements
> and modules. Traditional benchmark/leaderboard-driven bespoke approaches in
> the Machine Learning community are not suited to meet the highly demanding
> industrial standards beyond algorithmic performance, such as
> cost-effectiveness, safety, scalability, and automatability, typically
> expected in production systems. In this talk, I discuss some technical
> issues toward addressing these challenges: 1) a theoretical framework for
> panoramic learning with all experiences; 2) optimization methods to best
> the effort for learning under such a principled framework; 3) compositional
> strategies for building production-grade ML programs from standard parts. I
> will present our recent work toward developing a standard model for
> Learning that unifies different special-purpose machine learning paradigms
> and algorithms, then a Bayesian blackbox optimization approach to Meta
> Learning in the space of hyperparameters, model architectures, and system
> configurations, and finally principles and designs of standardized software
> Legos that facilitate cost-effective building, training, and tuning of
> practical ML pipelines and systems.
>
> *Speaker Bio*: Eric P. Xing is the President of the Mohamed bin Zayed
> University of Artificial Intelligence, a Professor of Computer Science at
> Carnegie Mellon University, and the Founder and Chairman of Petuum Inc., a
> 2018 World Economic Forum Technology Pioneer company that builds
> standardized artificial intelligence development platform and operating
> system for broad and general industrial AI applications. He completed his
> PhD in Computer Science at UC Berkeley. His main research interests are the
> development of machine learning and statistical methodology; and
> composable, automatic, and scalable computational systems, for solving
> problems involving automated learning, reasoning, and decision-making in
> artificial, biological, and social systems. Prof. Xing currently serves or
> has served the following roles: associate editor of the Journal of the
> American Statistical Association (JASA), Annals of Applied Statistics
> (AOAS), and IEEE Journal of Pattern Analysis and Machine Intelligence
> (PAMI); action editor of the Machine Learning Journal (MLJ) and Journal of
> Machine Learning Research (JMLR); he is a board member of the International
> Machine Learning Society.
>
> *In Person: *GHC 8102
> *Zoom Link*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>
> Thanks,
> Asher Trockman
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220513/ac4149e9/attachment.html>

From ashert at cs.cmu.edu  Mon May 16 11:54:48 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Mon, 16 May 2022 11:54:48 -0400
Subject: [CMU AI Seminar] May 17 at 12pm (Zoom) -- Machel Reid (U. Tokyo) --
 Incorporating Text Editing in Natural Language Processing -- AI Seminar
 sponsored by Morgan Stanley
Message-ID: <CAMVmiv2+rx=CPppoB2U7=ArcUpb9MGefSb50w9tOXHpZvvax-Q@mail.gmail.com>

Dear all,

We look forward to seeing you *tomorrow, this Tuesday (5/17)* from
*1**2:00-1:00
PM (U.S. Eastern time)* for the next talk of our *CMU AI seminar*,
sponsored by Morgan Stanley
<https://www.morganstanley.com/about-us/technology/>.

To learn more about the seminar series or see the future schedule, please
visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

*Tomorrow* (5/17), *Machel Reid *(U. Tokyo) will be giving a talk
titled *"**Incorporating Text
Editing in Natural Language Processing**".*

*Title*: Incorporating Text Editing in Natural Language Processing

*Talk Abstract*: Most current text generation applications in NLP use fully
autoregressive modelling, a notable example of this being the models in the
GPT series. However, most original content (e.g. art, books, articles,
source code) is developed not in a single iteration, but in many iterations
with each more refined than the last. In other words, revising and editing
are a central part of the human creative workflow. Given this, in this talk
I will talk about ideas for trying to bridge this disconnect by
incorporating text editing into standard natural language processing as
well as my work associated with various aspects of editing.

*Speaker Bio*: Machel Reid is a researcher at the University of Tokyo
working on natural language processing. He has worked on multilinguality,
primarily focusing on low-resource languages and multilinguality
pre-training. He is currently under the supervision of Yutaka Matsuo, and
has been an intern at Carnegie Mellon University under the supervision of
Graham Neubig. He is also an incoming PhD Student at the University of
Washington advised by Luke Zettlemoyer and Noah Smith.

*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220516/55478d0c/attachment.html>

From ashert at andrew.cmu.edu  Tue May 17 11:30:33 2022
From: ashert at andrew.cmu.edu (Asher Trockman)
Date: Tue, 17 May 2022 17:30:33 +0200
Subject: [CMU AI Seminar] May 17 at 12pm (Zoom) -- Machel Reid (U. Tokyo)
 -- Incorporating Text Editing in Natural Language Processing -- AI Seminar
 sponsored by Morgan Stanley
In-Reply-To: <CAMVmiv2+rx=CPppoB2U7=ArcUpb9MGefSb50w9tOXHpZvvax-Q@mail.gmail.com>
References: <CAMVmiv2+rx=CPppoB2U7=ArcUpb9MGefSb50w9tOXHpZvvax-Q@mail.gmail.com>
Message-ID: <1DB51C27-CA52-4DCB-962C-D400B86F7053@andrew.cmu.edu>

Hi all,

Just a reminder that Machel be will be giving a talk today at noon. Zoom link: https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher

> On May 16, 2022, at 5:54 PM, Asher Trockman <ashert at cs.cmu.edu> wrote:
> ?
> Dear all,
> 
> We look forward to seeing you tomorrow, this Tuesday (5/17) from 12:00-1:00 PM (U.S. Eastern time) for the next talk of our CMU AI seminar, sponsored by Morgan Stanley.
> 
> To learn more about the seminar series or see the future schedule, please visit the seminar website.
> 
> Tomorrow (5/17), Machel Reid (U. Tokyo) will be giving a talk titled "Incorporating Text Editing in Natural Language Processing".
> 
> Title: Incorporating Text Editing in Natural Language Processing
> 
> Talk Abstract: Most current text generation applications in NLP use fully autoregressive modelling, a notable example of this being the models in the GPT series. However, most original content (e.g. art, books, articles, source code) is developed not in a single iteration, but in many iterations with each more refined than the last. In other words, revising and editing are a central part of the human creative workflow. Given this, in this talk I will talk about ideas for trying to bridge this disconnect by incorporating text editing into standard natural language processing as well as my work associated with various aspects of editing.
> 
> Speaker Bio: Machel Reid is a researcher at the University of Tokyo working on natural language processing. He has worked on multilinguality, primarily focusing on low-resource languages and multilinguality pre-training. He is currently under the supervision of Yutaka Matsuo, and has been an intern at Carnegie Mellon University under the supervision of Graham Neubig. He is also an incoming PhD Student at the University of Washington advised by Luke Zettlemoyer and Noah Smith.
> 
> Zoom Link:  https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
> 
> Thanks,
> Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220517/9b3e8b85/attachment.html>

From ashert at cs.cmu.edu  Sun Sep 11 16:08:02 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Sun, 11 Sep 2022 16:08:02 -0400
Subject: [CMU AI Seminar] September 13 at 12pm (NSH 3305 & Zoom) -- Zico
 Kolter (CMU) -- New approaches to detecting and adapting to domain shifts in
 machine learning -- AI Seminar sponsored by SambaNova Systems
Message-ID: <CAMVmiv2YZZXSE2srg_tKswnYwnWqtU=yGjePu3cZe12dF8vCbg@mail.gmail.com>

Dear all,

We look forward to seeing you *this Tuesday (9/13)* from *1**2:00-1:00 PM
(U.S. Eastern time)* for the first talk of this semester's *CMU AI Seminar*,
sponsored by SambaNova Systems <https://sambanova.ai/>. The seminar will be
held in person in NSH 3305 with food provided, and it will also be
available on Zoom.

To learn more about the seminar series or to see the future schedule,
please visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

On 9/13, *Zico Kolter *(CMU) will be giving a talk titled *"New approaches
to detecting and adapting to domain shifts in machine learning**" *to share
work on evaluating and adapting machine learning models under distribution
shift.

*Title*: New approaches to detecting and adapting to domain shifts in
machine learning

*Talk Abstract*: Machine learning systems, in virtually every deployed
system, encounter data from a qualitatively different distribution than
what they were trained upon. Effectively dealing with this problem, known
as domain shift, is thus perhaps the key challenge in deploying machine
learning methods in practice. In this talk, I will motivate some of these
challenges in domain shift, and highlight some of our recent work on two
topics. First, I will present our work on determining if we can even just
evaluate the performance of machine learning models under distribution
shift, without access to labelled data. And second, I will present work on
how we can better adapt our classifiers to new data distributions, again
assuming access only to unlabelled data in the new domain.

*Speaker Bio*: Zico Kolter <http://zicokolter.com> is an Associate
Professor in the Computer Science Department at Carnegie Mellon University,
and also serves as chief scientist of AI research for the Bosch Center for
Artificial Intelligence. His work spans the intersection of machine
learning and optimization, with a large focus on developing more robust and
rigorous methods in deep learning. In addition, he has worked in a number
of application areas, highlighted by work on sustainability and smart
energy systems. He is a recipient of the DARPA Young Faculty Award, a Sloan
Fellowship, and best paper awards at NeurIPS, ICML (honorable mention),
IJCAI, KDD, and PESGM.

*In person: *NSH 3305
*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220911/41b96189/attachment.html>

From ashert at cs.cmu.edu  Tue Sep 13 11:14:35 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 13 Sep 2022 11:14:35 -0400
Subject: [CMU AI Seminar] September 13 at 12pm (NSH 3305 & Zoom) -- Zico
 Kolter (CMU) -- New approaches to detecting and adapting to domain shifts in
 machine learning -- AI Seminar sponsored by SambaNova Systems
In-Reply-To: <CAMVmiv2YZZXSE2srg_tKswnYwnWqtU=yGjePu3cZe12dF8vCbg@mail.gmail.com>
References: <CAMVmiv2YZZXSE2srg_tKswnYwnWqtU=yGjePu3cZe12dF8vCbg@mail.gmail.com>
Message-ID: <CAMVmiv3RU1BU-u1HL1FqvL98cGYNgMAtj5H+6OypdCfhGQ8eJg@mail.gmail.com>

Hi all,

Reminder that this is happening today at noon in NSH 3305 (and on Zoom).
There will be pizza. Hope to see you there!

Thanks,
Asher

On Sun, Sep 11, 2022 at 4:08 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Dear all,
>
> We look forward to seeing you *this Tuesday (9/13)* from *1**2:00-1:00 PM
> (U.S. Eastern time)* for the first talk of this semester's
> *CMU AI Seminar*, sponsored by SambaNova Systems <https://sambanova.ai/>.
> The seminar will be held in person in NSH 3305 with food provided, and it
> will also be available on Zoom.
>
> To learn more about the seminar series or to see the future schedule,
> please visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>
> On 9/13, *Zico Kolter *(CMU) will be giving a talk titled *"New
> approaches to detecting and adapting to domain shifts in machine learning**"
> *to share work on evaluating and adapting machine learning models under
> distribution shift.
>
> *Title*: New approaches to detecting and adapting to domain shifts in
> machine learning
>
> *Talk Abstract*: Machine learning systems, in virtually every deployed
> system, encounter data from a qualitatively different distribution than
> what they were trained upon. Effectively dealing with this problem, known
> as domain shift, is thus perhaps the key challenge in deploying machine
> learning methods in practice. In this talk, I will motivate some of these
> challenges in domain shift, and highlight some of our recent work on two
> topics. First, I will present our work on determining if we can even just
> evaluate the performance of machine learning models under distribution
> shift, without access to labelled data. And second, I will present work on
> how we can better adapt our classifiers to new data distributions, again
> assuming access only to unlabelled data in the new domain.
>
> *Speaker Bio*: Zico Kolter <http://zicokolter.com> is an Associate
> Professor in the Computer Science Department at Carnegie Mellon University,
> and also serves as chief scientist of AI research for the Bosch Center for
> Artificial Intelligence. His work spans the intersection of machine
> learning and optimization, with a large focus on developing more robust and
> rigorous methods in deep learning. In addition, he has worked in a number
> of application areas, highlighted by work on sustainability and smart
> energy systems. He is a recipient of the DARPA Young Faculty Award, a Sloan
> Fellowship, and best paper awards at NeurIPS, ICML (honorable mention),
> IJCAI, KDD, and PESGM.
>
> *In person: *NSH 3305
> *Zoom Link*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>
> Thanks,
> Asher Trockman
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220913/ae0dd658/attachment.html>

From ashert at cs.cmu.edu  Tue Sep 13 19:26:10 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 13 Sep 2022 19:26:10 -0400
Subject: [CMU AI Seminar] Special! September 14 at 2pm (NSH 4305 & Virtually)
 -- Steve Chien (JPL) -- AI in Space - From Earth Orbit to Mars and
 Beyond! -- AI Seminar sponsored by SambaNova Systems
Message-ID: <CAMVmiv1xDWihkNb8JbZ2Z+nsWnFDA-fhEEGmf_14QLOoBu+Qpw@mail.gmail.com>

Dear all,

We invite you to a special installment of our CMU AI Seminar Series* tomorrow,
this Wednesday (9/14)* in *NSH 4305 *and on WebEx (link below) from *2:00 -
3:00 PM (U.S. Eastern time)*, sponsored by SambaNova Systems
<https://sambanova.ai/>.

To learn more about the seminar series or to see the future schedule,
please visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

Tomorrow (Wed. 9/14), *Steve Chien *(JPL) will be giving a talk titled *"AI
in Space -- From Earth Orbit to Mars and Beyond!**".*

*Title*: AI in Space -- From Earth Orbit to Mars and Beyond!

*Abstract: *Artificial Intelligence is playing an increasing role in our
everyday lives and the business marketplace. This trend extends to the
space sector, where AI has already shown considerable success and has the
potential to revolutionize almost every aspect of space exploration. We
first highlight a number of success stories of the tremendous impact of
Artificial Intelligence in Space: over a dozen years of operations of the
Autonomous Sciencecraft on EO-1, the Earth Observing Sensorweb tracking
volcanoes, flooding and wildfires and automated targeting onboard the MER
and MSL rovers. Finally we discuss why AI is essential to the search for
life beyond Earth, highlighting the key role of AI in Europa Submersible
and Interstellar mission concepts.

*Speaker Bio*: Dr. Steve Chien
<https://ai.jpl.nasa.gov/public/people/chien/> is JPL Fellow, Senior
Research Scientist, and Technical Group Supervisor of the Artificial
Intelligence Group and in the Mission Planning and Execution Section at the
Jet Propulsion Laboratory, California Institute of Technology where he
leads efforts in automated planning and scheduling for space exploration.
Dr. Chien was previously Adjunct Faculty with the Department of Computer
Science of the University of Southern California, and a Research Scientist
at the Joint Institute for Regional Earth System Science & Engineering and
a Visiting Scholar with the Department of Computer Science of the
University of California at Los Angeles. He holds a B.S. with Highest
Honors in Computer Science, with minors in Mathematics and Economics, M.S.,
and Ph.D. degrees in Computer Science, all from the University of Illinois.

*In person: *NSH 4305
*WebEx: *
https://jpl.webex.com/jpl/j.php?MTID=mdc1f706256a769341352bfa976bfeb20
(You can join in your browser.)

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220913/d6d98562/attachment.html>

From ashert at cs.cmu.edu  Sun Sep 25 17:03:02 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Sun, 25 Sep 2022 17:03:02 -0400
Subject: [CMU AI Seminar] September 27 at 12pm (NSH 3305 & Zoom) -- Jian Zhang
 (SambaNova Systems) -- MLSys Innovations Beyond the Hardware
 Goldilocks Zone -- AI Seminar sponsored by SambaNova Systems
Message-ID: <CAMVmiv2Et5mQ=tjT_gdZT8KbvtE2mKES+59er5YBQQtK6YUYbA@mail.gmail.com>

Dear all,

We look forward to seeing you *this Tuesday (9/27)* from *1**2:00-1:00 PM
(U.S. Eastern time)* for the next talk of this semester's *CMU AI Seminar*,
sponsored by SambaNova Systems <https://sambanova.ai/>. The seminar will be
held in NSH 3305 *with pizza provided *and will be streamed on Zoom.

To learn more about the seminar series or to see the future schedule,
please visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

On 9/27, *Jian Zhang* (SambaNova Systems) will be giving a talk titled
*"**MLSys
Innovations Beyond the Hardware Goldilocks Zone**" *to share
SambaNova's recent advances in software/hardware co-design for sparse
training (e.g., of large language models).

*Title*: MLSys Innovations Beyond the Hardware Goldilocks Zone

*Talk Abstract*: The Goldilocks zone, or habitable zone, limits life to a
restricted part of the solar system. In a similar way, the limitations of
conventional hardware impose a Goldilocks zone for ML system innovations.
In recent years, the emerging deep learning accelerators have launched an
unprecedented opportunity for ML systems advance at different layers of the
software and machine learning stacks. In this tech talk, I am very excited
to share a case study for building large scale models on the reconfigurable
dataflow units (RDU) at SambaNova Systems. At the ML application layer, I
want to highlight the 0-1 accuracy breakthrough in High Resolution 3D
segmentation enabled by the large memory capacity in RDU. In the layer of
ML training algorithms, I will unbox our recent advance in SW/HW codesign
for RDU sparse training; this endeavor leads to a 6X faster
time-to-accuracy for pretraining large language models over standard dense
training on A100 GPUs. Lastly, I will showcase how our team built the
prototype of learned RDU performance optimization methods, which pushes
further towards fully unleashing the hardware capability of RDU with
significantly reduced engineering cost than rule-based methods. We hope
that the emerging accelerators will trigger many more ML system innovations
from the community.

*Speaker Bio*: Jian Zhang is the Director of Machine Learning at SambaNova
Systems. He leads the ML team which builds the deep learning foundations
for SambaNova?s large-scale enterprise AI solutions. With the mission of
democratizing modern foundation model systems, the ML team at SambaNova
innovates on both the machine learning and the system aspects,
including productionalizing large foundation models and ML/hardware
co-design on emerging hardware. Before joining SambaNova Systems, Jian got
his PhD in machine learning from Stanford University focusing on machine
learning and natural language processing systems.

*In person: *NSH 3305
*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220925/3318de5d/attachment.html>

From ashert at andrew.cmu.edu  Tue Sep 27 11:58:37 2022
From: ashert at andrew.cmu.edu (Asher Trockman)
Date: Tue, 27 Sep 2022 11:58:37 -0400
Subject: [CMU AI Seminar] September 27 at 12pm (NSH 3305 & Zoom) -- Jian
 Zhang (SambaNova Systems) -- MLSys Innovations Beyond the Hardware Goldilocks
 Zone -- AI Seminar sponsored by SambaNova Systems
In-Reply-To: <CAMVmiv2Et5mQ=tjT_gdZT8KbvtE2mKES+59er5YBQQtK6YUYbA@mail.gmail.com>
References: <CAMVmiv2Et5mQ=tjT_gdZT8KbvtE2mKES+59er5YBQQtK6YUYbA@mail.gmail.com>
Message-ID: <D1D43910-E518-4671-9FF3-5D9E8EA8B7E0@andrew.cmu.edu>

This is happening now in NSH 3305 (there's pizza).

> On Sep 25, 2022, at 5:03 PM, Asher Trockman <ashert at cs.cmu.edu> wrote:
> 
> ?
> Dear all,
> 
> We look forward to seeing you this Tuesday (9/27) from 12:00-1:00 PM (U.S. Eastern time) for the next talk of this semester's CMU AI Seminar, sponsored by SambaNova Systems. The seminar will be held in NSH 3305 with pizza provided and will be streamed on Zoom.
> 
> To learn more about the seminar series or to see the future schedule, please visit the seminar website.
> 
> On 9/27, Jian Zhang (SambaNova Systems) will be giving a talk titled "MLSys Innovations Beyond the Hardware Goldilocks Zone" to share SambaNova's recent advances in software/hardware co-design for sparse training (e.g., of large language models).
> 
> Title: MLSys Innovations Beyond the Hardware Goldilocks Zone
> 
> Talk Abstract: The Goldilocks zone, or habitable zone, limits life to a restricted part of the solar system. In a similar way, the limitations of conventional hardware impose a Goldilocks zone for ML system innovations. In recent years, the emerging deep learning accelerators have launched an unprecedented opportunity for ML systems advance at different layers of the software and machine learning stacks. In this tech talk, I am very excited to share a case study for building large scale models on the reconfigurable dataflow units (RDU) at SambaNova Systems. At the ML application layer, I want to highlight the 0-1 accuracy breakthrough in High Resolution 3D segmentation enabled by the large memory capacity in RDU. In the layer of ML training algorithms, I will unbox our recent advance in SW/HW codesign for RDU sparse training; this endeavor leads to a 6X faster time-to-accuracy for pretraining large language models over standard dense training on A100 GPUs. Lastly, I will showcase how our team built the prototype of learned RDU performance optimization methods, which pushes further towards fully unleashing the hardware capability of RDU with significantly reduced engineering cost than rule-based methods. We hope that the emerging accelerators will trigger many more ML system innovations from the community.
> 
> Speaker Bio: Jian Zhang is the Director of Machine Learning at SambaNova Systems. He leads the ML team which builds the deep learning foundations for SambaNova?s large-scale enterprise AI solutions. With the mission of democratizing modern foundation model systems, the ML team at SambaNova innovates on both the machine learning and the system aspects, including productionalizing large foundation models and ML/hardware co-design on emerging hardware. Before joining SambaNova Systems, Jian got his PhD in machine learning from Stanford University focusing on machine learning and natural language processing systems.
> 
> In person: NSH 3305
> Zoom Link:  https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
> 
> Thanks,
> Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20220927/405419c5/attachment.html>

From ashert at cs.cmu.edu  Mon Oct 24 15:58:05 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Mon, 24 Oct 2022 15:58:05 -0400
Subject: [CMU AI Seminar] October 25 at 12pm (NSH 3305 & Zoom) -- Charvi
 Rastogi (CMU) -- Two Studies on Peer Review: Finding implicit biases in
 conference peer review -- AI Seminar sponsored by SambaNova Systems
Message-ID: <CAMVmiv2F2jCpkqQ-ZYH4R1cP-XRJ_gu5nkqYM4qgz8YrcekkVQ@mail.gmail.com>

Dear all,

We look forward to seeing you *tomorrow, **this Tuesday (10/25)* from
*1**2:00-1:00
PM (U.S. Eastern time)* for the next talk of this semester's
*CMU AI Seminar*, sponsored by SambaNova Systems <https://sambanova.ai/>.
The seminar will be held in NSH 3305 *with pizza provided *and will be
streamed on Zoom.

To learn more about the seminar series or to see the future schedule,
please visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

Tomorrow (10/25), *Charvi Rastogi* (CMU MLD) will be giving a talk titled
*"**Two Studies on Peer Review: Finding implicit biases in conference peer
review**".*

*Title*: Two Studies on Peer Review: Finding implicit biases in conference
peer review

*Talk Abstract*: Modern peer review systems are a huge part of academia,
and play a consequential role in participants' career trajectories. In this
talk, we will discuss two studies (involving several thousand papers and
reviewers) that scrutinize certain aspects of the peer-review process in
two conferences: ICML and EC. In this work, we focus on sources of implicit
bias in the decision-making process. In the first study, we investigate the
positive and negative effects of posting preprints online for multiple
stakeholders. ArXiv is an increasingly popular choice for posting
preprints, however it stands to dilute anonymity in double-blind peer
review. This can negatively impact the review process for authors
affiliated to lower ranked institutions. We conduct a study to help
quantify the associated risks and benefits of posting preprints online for
authors from different institutions. In the second study, we investigate
whether cited reviewers are implicitly biased in favor of acceptance of the
paper. For this, we intervene in the paper-reviewer assignment procedure
and carefully analyze the outcomes of the review process (observational
data) to test for the presence of citation bias. Our work has  policy
implications for conference peer-review design and authors' engagement with
it.

*In person: *NSH 3305
*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20221024/a72d1db6/attachment.html>

From ashert at cs.cmu.edu  Tue Oct 25 11:28:53 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 25 Oct 2022 11:28:53 -0400
Subject: [CMU AI Seminar] Special! October 26 at 1:30pm (NSH 3305 & Zoom) --
 John P. Dickerson (UMD) -- Robustness, Privacy, Fairness, and Credibility?
 Pushing the Boundaries of Economic Design with Deep Learning -- AI Seminar
 sponsored by SambaNova Systems
Message-ID: <CAMVmiv2E1ULULQ07fjNMM-1iiOfGgogitkD2njLZ0E1cTm0eiw@mail.gmail.com>

Dear all,

We invite you to a special installment of our CMU AI Seminar Series* tomorrow,
this Wednesday (10/26)* from *1:30 - 2:30** PM (U.S. Eastern time)*,
sponsored by SambaNova Systems <https://sambanova.ai/>. The seminar will be
held in NSH 3305 *with pizza provided *and will be streamed on Zoom.

To learn more about the seminar series or to see the future schedule,
please visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

Tomorrow (10/26), *John P. Dickerson* (UMD) will be giving a talk
titled *"**Robustness,
Privacy, Fairness, and Credibility? Pushing the Boundaries of Economic
Design with Deep Learning**".*

*Title*: Robustness, Privacy, Fairness, and Credibility? Pushing the
Boundaries of Economic Design with Deep Learning

*Talk Abstract*: The design of revenue-maximizing auctions with strong
incentive guarantees is a core concern of economic theory. Computational
auctions enable online advertising, sourcing, spectrum allocation, and
myriad financial markets. Analytic progress in this space is notoriously
difficult; since Myerson's 1981 work characterizing single-item optimal
auctions, there has been limited progress outside of restricted settings. A
recent paper by D?tting et al. circumvents analytic difficulties by
applying deep learning techniques to, instead, approximate optimal
auctions. Their RegretNet architecture can represent auctions with
arbitrary numbers of items and participants; it is trained to be
empirically strategyproof, but the property is never exactly verified
leaving potential loopholes for market participants to exploit. In
parallel, new research from Ilvento et al. and other groups has developed
notions of fairness in the context of auction design. Inspired by these
advances, in this talk, we discuss extensions of these techniques for
approximating auctions using deep learning to address concerns of
* fairness while maintaining high revenue and strong incentive guarantees,
including learning fairness from human preferences;
* certified robustness, that is, verification of claimed strategyproofness
of deep learned auctions; and
* expressiveness via different demand functions and other constraints.

To enable that last point, we propose a new architecture to learn incentive
compatible, revenue-maximizing auctions from sampled valuations, which uses
the Sinkhorn algorithm to perform a differentiable bipartite matching. Our
new framework allows the network to learn strategyproof revenue-maximizing
mechanisms in settings not learnable by the previous RegretNet
architecture.  This talk connects work in the deep learning for auction
design space into the deep learning for matching market design space, and
provides concrete steps forward regarding differentiable economics and
matching market design.

*Speaker Bio: *John P Dickerson is co-founder and Chief Scientist of
Arthur, the AI performance monitoring company, as well as Associate
Professor of Computer Science at the University of Maryland.  He is a
recipient of awards such as the NSF CAREER Award, IEEE Intelligent Systems
AI's 10 to Watch, Google Faculty Research Award, Google AI for Social Good
Award, and paper awards and nominations at venues such as AAAI.  His
research centers on solving practical economic problems using techniques
from computer science, stochastic optimization, and machine learning. He
has worked extensively on theoretical and empirical approaches to organ
exchange where his work has set policy at the UNOS nationwide kidney
exchange; worldwide blood donation markets with Facebook; game-theoretic
approaches to counter-terrorism and negotiation, where his models have been
deployed; and market design problems in industry (e.g., online advertising)
through various startups. He received his PhD in computer science from
Carnegie Mellon University (SCS CSD PhD '16).

*In person: *NSH 3305
*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20221025/a3aa7d40/attachment.html>

From ashert at cs.cmu.edu  Sun Oct 30 11:49:07 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Sun, 30 Oct 2022 11:49:07 -0400
Subject: =?UTF-8?Q?=5BCMU_AI_Seminar=5D_November_1_at_12pm_=28NSH_3305_=26_Zoom?=
 =?UTF-8?Q?=29_=2D=2D_S=C3=A9bastien_Bubeck_=28Microsoft_Research=29_=2D=2D_Unveiling_Tra?=
 =?UTF-8?Q?nsformers_with_LEGO_=2D=2D_AI_Seminar_sponsored_by_SambaNova_Sys?=
 =?UTF-8?Q?tems?=
Message-ID: <CAMVmiv20t__RvaqMtvcs=qig+QpcipiKhgffoNsooyPBdgmqxA@mail.gmail.com>

Dear all,

We look forward to seeing you *this Tuesday (11/1)* from *1**2:00-1:00 PM
(U.S. Eastern time)* for the next talk of this semester's *CMU AI Seminar*,
sponsored by SambaNova Systems <https://sambanova.ai/>. The seminar will be
held in NSH 3305 *with pizza provided *and will be streamed on Zoom.

To learn more about the seminar series or to see the future schedule,
please visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

On 11/1, *S?bastien Bubeck* (Microsoft Research) will be giving a talk
titled *"**Unveiling Transformers with LEGO**" *to share his recent work on
probing the inner workings of transformers using a synthetic reasoning task.

*Title*: Unveiling Transformers with LEGO

*Talk Abstract*: The discovery of the transformer architecture was a
paradigm shifting event for deep learning. However, these architectures are
arguably even harder to understand than say convolutional neural networks.
In this work we propose a synthetic reasoning task, called LEGO, to probe
the inner workings of transformers. We obtain some insights on multi-head
attention, the effect of pretraining, as well as overfitting issues.
Joint work with Yi Zhang, Arturs Backurs, Ronen Eldan, Suriya Gunasekar,
and Tal Wagner.

*Speaker Bio*: Sebastien Bubeck leads the Machine Learning Foundations
group at Microsoft Research Redmond. He joined MSR in 2014, after three
years as an assistant professor at Princeton University. He received
several best paper awards at machine learning conferences for his work on
online decision making, convex optimization, and adversarial robustness
(NeurIPS 2021, NeurIPS 2018, ALT 2018, COLT 2016, COLT 2009). He also wrote
two monographs, ?Regret Analysis of Stochastic and Non-Stochastic
Multi-Armed Bandit Problems? (2012) and ?Convex Optimization: Algorithms
and Complexity? (2014).

*In person: *NSH 3305
*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20221030/3d60386c/attachment.html>

From ashert at cs.cmu.edu  Fri Nov  4 16:17:10 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Fri, 4 Nov 2022 16:17:10 -0400
Subject: [CMU AI Seminar] November 8 at 12pm (NSH 3305 & Zoom) -- Aditi
 Raghunathan (CMU) -- Robustness in the era of large pretrained models -- AI
 Seminar sponsored by SambaNova Systems
Message-ID: <CAMVmiv18t3Y7z4YvVD-=eqJg=1THur4XNF8oSeKT0E8T+GP7HQ@mail.gmail.com>

Dear all,

We look forward to seeing you *this coming Tuesday (11/8)* from *1**2:00-1:00
PM (U.S. Eastern time)* for the next talk of this semester's
*CMU AI Seminar*, sponsored by SambaNova Systems <https://sambanova.ai/>.
The seminar will be held in NSH 3305 *with pizza provided *and will be
streamed on Zoom.

To learn more about the seminar series or to see the future schedule,
please visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

On 11/8, *Aditi Raghunathan* (CMU) will be giving a talk titled *"**Robustness
in the era of large pretrained models**".*

*Title*: Robustness in the era of large pretrained models

*Talk Abstract*: Machine learning systems often fail catastrophically under
the presence of distribution shift?when the test distribution differs in
some systematic way from the training distribution. This notion of
robustness has remained an open challenge. The past few years have seen the
rise of large models trained on broad data at scale that can be adapted to
several downstream tasks (e.g. BERT, GPT, DALL-E). In this talk, via theory
and experiments, we will discuss how such models open up new avenues, but
also require new techniques for improving robustness.

*Speaker Bio*: Aditi Raghunathan is an assistant professor of computer
science at Carnegie Mellon University. She is interested in building robust
ML systems with guarantees for trustworthy real-world deployment.
Previously, she was a postdoctoral researcher at Berkeley AI Research, and
received her PhD from Stanford University in 2021. Her research has been
recognized by the Arthur Samuel Best Thesis Award at Stanford, a Google PhD
fellowship in machine learning, and an Open Philanthropy AI fellowship.

*In person: *NSH 3305
*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20221104/39bf8889/attachment.html>

From ashert at cs.cmu.edu  Fri Nov 11 16:59:12 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Fri, 11 Nov 2022 16:59:12 -0500
Subject: [CMU AI Seminar] November 15 at 12pm (NSH 3305 & Zoom) -- Alexander
 Terenin (University of Cambridge) -- Pathwise Conditioning and Non-Euclidean
 Gaussian Processes -- AI Seminar sponsored by SambaNova Systems
Message-ID: <CAMVmiv2dWg3=g_YMUJeUHUWPvEBcC_Z31siVqDp35J5-FAAuGg@mail.gmail.com>

Dear all,

We look forward to seeing you *this coming Tuesday (11/15)* from *1**2:00-1:00
PM (U.S. Eastern time)* for the next talk of this semester's
*CMU AI Seminar*, sponsored by SambaNova Systems <https://sambanova.ai/>.
The seminar will be held in NSH 3305 *with pizza provided *and will be
streamed on Zoom.

To learn more about the seminar series or to see the future schedule,
please visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

On 11/15, *Alexander Terenin* (University of Cambridge) will be giving a
talk titled *"**Pathwise Conditioning and Non-Euclidean Gaussian Processes*
*".*

*Title*: Pathwise Conditioning and Non-Euclidean Gaussian Processes

*Talk Abstract*: In Gaussian processes, conditioning and computation of
posterior distributions is usually done in a distributional fashion by
working with finite-dimensional marginals. However, there is another way to
think about conditioning: using actual random functions rather than their
probability distributions. This perspective is particularly helpful in
decision-theoretic settings such as Bayesian optimization, where it enables
efficient computation of a wider class of acquisition functions than
otherwise possible. In this talk, we describe these recent advances, and
discuss their broader implications to Gaussian processes. We then present a
class of Gaussian process models on graphs and manifolds, which can enable
one to perform Bayesian optimization while taking into account symmetries
and constraints in an intrinsic manner.

*Speaker Bio*: Alexander Terenin is a Postdoctoral Research Associate at
the University of Cambridge. He is interested in statistical machine
learning, particularly in settings where the data is not fixed, but is
gathered interactively by the learning machine. This leads naturally to
Gaussian processes and data-efficient interactive decision-making systems
such as Bayesian optimization, to areas such as multi-armed bandits and
reinforcement learning, and to techniques for incorporating inductive
biases and prior information such as symmetries into machine learning
models.

*In person: *NSH 3305
*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20221111/c0884d71/attachment.html>

From ashert at cs.cmu.edu  Tue Nov 15 10:08:45 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 15 Nov 2022 10:08:45 -0500
Subject: [CMU AI Seminar] November 15 at 12pm (NSH 3305 & Zoom) --
 Alexander Terenin (University of Cambridge) -- Pathwise Conditioning and
 Non-Euclidean Gaussian Processes -- AI Seminar sponsored by SambaNova Systems
In-Reply-To: <CAMVmiv2dWg3=g_YMUJeUHUWPvEBcC_Z31siVqDp35J5-FAAuGg@mail.gmail.com>
References: <CAMVmiv2dWg3=g_YMUJeUHUWPvEBcC_Z31siVqDp35J5-FAAuGg@mail.gmail.com>
Message-ID: <CAMVmiv0cY0Ojo-foM_yH6zgu0SG20XNfypyoEg11Xe70QJys1A@mail.gmail.com>

Just a reminder that this is happening today.

On Fri, Nov 11, 2022 at 4:59 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Dear all,
>
> We look forward to seeing you *this coming Tuesday (11/15)* from *1**2:00-1:00
> PM (U.S. Eastern time)* for the next talk of this semester's
> *CMU AI Seminar*, sponsored by SambaNova Systems <https://sambanova.ai/>.
> The seminar will be held in NSH 3305 *with pizza provided *and will be
> streamed on Zoom.
>
> To learn more about the seminar series or to see the future schedule,
> please visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>
> On 11/15, *Alexander Terenin* (University of Cambridge) will be giving a
> talk titled *"**Pathwise Conditioning and Non-Euclidean Gaussian
> Processes**".*
>
> *Title*: Pathwise Conditioning and Non-Euclidean Gaussian Processes
>
> *Talk Abstract*: In Gaussian processes, conditioning and computation of
> posterior distributions is usually done in a distributional fashion by
> working with finite-dimensional marginals. However, there is another way to
> think about conditioning: using actual random functions rather than their
> probability distributions. This perspective is particularly helpful in
> decision-theoretic settings such as Bayesian optimization, where it enables
> efficient computation of a wider class of acquisition functions than
> otherwise possible. In this talk, we describe these recent advances, and
> discuss their broader implications to Gaussian processes. We then present a
> class of Gaussian process models on graphs and manifolds, which can enable
> one to perform Bayesian optimization while taking into account symmetries
> and constraints in an intrinsic manner.
>
> *Speaker Bio*: Alexander Terenin is a Postdoctoral Research Associate at
> the University of Cambridge. He is interested in statistical machine
> learning, particularly in settings where the data is not fixed, but is
> gathered interactively by the learning machine. This leads naturally to
> Gaussian processes and data-efficient interactive decision-making systems
> such as Bayesian optimization, to areas such as multi-armed bandits and
> reinforcement learning, and to techniques for incorporating inductive
> biases and prior information such as symmetries into machine learning
> models.
>
> *In person: *NSH 3305
> *Zoom Link*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>
> Thanks,
> Asher Trockman
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20221115/38fc48d4/attachment.html>

From ashert at cs.cmu.edu  Sun Nov 20 13:44:55 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Sun, 20 Nov 2022 13:44:55 -0500
Subject: [CMU AI Seminar] November 22 at 12pm (NSH 3305 & Zoom) -- Aldo
 Pacchiano (Microsoft Research) -- Online Model Selection: the principle of
 regret balancing -- AI Seminar sponsored by SambaNova Systems
Message-ID: <CAMVmiv2FucOva11DgByxR5_vec8UPoS4Xgp=6UWQU_DFjj9FSw@mail.gmail.com>

Dear all,

We look forward to seeing you *this coming Tuesday (11/22)* from *1**2:00-1:00
PM (U.S. Eastern time)* for the next talk of this semester's
*CMU AI Seminar*, sponsored by SambaNova Systems <https://sambanova.ai/>.
The seminar will be held in NSH 3305 *with pizza provided *and will be
streamed on Zoom.

To learn more about the seminar series or to see the future schedule,
please visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.

On 11/22, *Aldo Pacchiano* (Microsoft Research) will be giving a talk
titled *"**Online Model Selection: the principle of regret balancing**".*

*Title*: Online Model Selection: the principle of regret balancing

*Talk Abstract*: We will introduce the problem of online model selection
where a learner is to select among a set of online algorithms to solve a
specific problem instance. We would like to design algorithms that allow
such a learner to select in an online fashion the best algorithm without
incurring much regret. This problem is challenging because in contrast with
for example multi armed bandits, the algorithms' rewards -due to the
algorithm's own learning process- may be non-stationary. We will introduce
the principle of regret balancing, a simple, practical and effective model
selection algorithmic design technique that allows for online selection of
the best among multiple (base) algorithms in a fully blackbox fashion.
Regret balancing solves the problem of non-stationarity by introducing an
elegant `misspecification test' that can efficiently detect when a base
algorithm is not appropriate for the problem at hand. Regret balancing
techniques have also been used to provide clarity to some long-standing
problems in online learning such as corruption learning in MDPs.

*Speaker Bio*: Aldo Pacchiano is a postdoctoral researcher at Microsoft
Research NYC. He obtained his PhD at UC Berkeley where he was advised by
Prof. Peter Bartlett and Prof. Michael Jordan. His research lies in the
areas of Reinforcement Learning, Online Learning, Bandits and Algorithmic
Fairness. He is particularly interested in furthering our statistical
understanding of learning phenomena in adaptive environments and use these
theoretical insights and techniques to design useful algorithms in (among
other things) bandits, RL, and experimental design.

*In person: *NSH 3305
*Zoom Link*:
https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09

Thanks,
Asher Trockman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20221120/be66deda/attachment.html>

From ashert at cs.cmu.edu  Tue Nov 22 11:58:03 2022
From: ashert at cs.cmu.edu (Asher Trockman)
Date: Tue, 22 Nov 2022 11:58:03 -0500
Subject: [CMU AI Seminar] November 22 at 12pm (NSH 3305 & Zoom) -- Aldo
 Pacchiano (Microsoft Research) -- Online Model Selection: the principle of
 regret balancing -- AI Seminar sponsored by SambaNova Systems
In-Reply-To: <CAMVmiv2FucOva11DgByxR5_vec8UPoS4Xgp=6UWQU_DFjj9FSw@mail.gmail.com>
References: <CAMVmiv2FucOva11DgByxR5_vec8UPoS4Xgp=6UWQU_DFjj9FSw@mail.gmail.com>
Message-ID: <CAMVmiv1T3nexpQ9aggygvPfLF=MtRhD0nFcYrHzY6Et=eS7kqQ@mail.gmail.com>

Reminder that this is happening now.

On Sun, Nov 20, 2022 at 1:44 PM Asher Trockman <ashert at cs.cmu.edu> wrote:

> Dear all,
>
> We look forward to seeing you *this coming Tuesday (11/22)* from *1**2:00-1:00
> PM (U.S. Eastern time)* for the next talk of this semester's
> *CMU AI Seminar*, sponsored by SambaNova Systems <https://sambanova.ai/>.
> The seminar will be held in NSH 3305 *with pizza provided *and will be
> streamed on Zoom.
>
> To learn more about the seminar series or to see the future schedule,
> please visit the seminar website <http://www.cs.cmu.edu/~aiseminar/>.
>
> On 11/22, *Aldo Pacchiano* (Microsoft Research) will be giving a talk
> titled *"**Online Model Selection: the principle of regret balancing**".*
>
> *Title*: Online Model Selection: the principle of regret balancing
>
> *Talk Abstract*: We will introduce the problem of online model selection
> where a learner is to select among a set of online algorithms to solve a
> specific problem instance. We would like to design algorithms that allow
> such a learner to select in an online fashion the best algorithm without
> incurring much regret. This problem is challenging because in contrast with
> for example multi armed bandits, the algorithms' rewards -due to the
> algorithm's own learning process- may be non-stationary. We will introduce
> the principle of regret balancing, a simple, practical and effective model
> selection algorithmic design technique that allows for online selection of
> the best among multiple (base) algorithms in a fully blackbox fashion.
> Regret balancing solves the problem of non-stationarity by introducing an
> elegant `misspecification test' that can efficiently detect when a base
> algorithm is not appropriate for the problem at hand. Regret balancing
> techniques have also been used to provide clarity to some long-standing
> problems in online learning such as corruption learning in MDPs.
>
> *Speaker Bio*: Aldo Pacchiano is a postdoctoral researcher at Microsoft
> Research NYC. He obtained his PhD at UC Berkeley where he was advised by
> Prof. Peter Bartlett and Prof. Michael Jordan. His research lies in the
> areas of Reinforcement Learning, Online Learning, Bandits and Algorithmic
> Fairness. He is particularly interested in furthering our statistical
> understanding of learning phenomena in adaptive environments and use these
> theoretical insights and techniques to design useful algorithms in (among
> other things) bandits, RL, and experimental design.
>
> *In person: *NSH 3305
> *Zoom Link*:
> https://cmu.zoom.us/j/99510233317?pwd=ZGx4aExNZ1FNaGY4SHI3Qlh0YjNWUT09
>
> Thanks,
> Asher Trockman
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/ai-seminar-announce/attachments/20221122/dd4ae769/attachment.html>