From awd at cs.cmu.edu  Tue May 10 11:07:17 2022
From: awd at cs.cmu.edu (Artur Dubrawski)
Date: Tue, 10 May 2022 11:07:17 -0400
Subject: Fwd: Intern offer
In-Reply-To: <0c126ef1-ed59-3833-aecf-a2b75abb6ca2@cs.cmu.edu>
References: <0c126ef1-ed59-3833-aecf-a2b75abb6ca2@cs.cmu.edu>
Message-ID: <CAJvAoyvRxx3axTdtU-ATZOu4c27z5gJzGYUxFQW5r42u71_ZCw@mail.gmail.com>

I know we are going to be swamped with interns this summer, most of whom
would likely be more capable on the start than this one,
but if you'd like to contribute to outreach by mentoring a talented high
schooler with little if any AI experience, please let me know.

Artur

---------- Forwarded message ---------
From: Illah Nourbakhsh <illah at cs.cmu.edu>
Date: Tue, May 10, 2022 at 11:00 AM
Subject: Intern offer
To: <robotics-faculty at lists.andrew.cmu.edu>


Hello friends and faculty,

We are beginning a pilot experiment with CCAC this year to grow in
future years. They are identifying students from marginalized
backgrounds who are in high school and are taking CCAC's early college
courses and doing very well. CCAC will offer the student a stipend
($3.5K) and a summer course; and I have volunteered to help place the
student in a lab at CMU over the summer- 4 days a week of work for
approx. 12 weeks- so the student can learn about being on campus and
doing research. All costs are covered by CCAC and anything extra the
Center for Shared Prosperity will cover (like, a bus pass). We will also
place them with Tartan Scholars so they get some college strategizing.

Lillian Martin is the first such student. She's as Oakland Catholic; I
have talked to her mother as well and they're extremely excited about
this prospect. Her interests are listed below; she just did the CS AP
test and knows Python and Javascript. If you could consider hosting her
for the summer, please email me. This is part of our effort to come up
with structural ways of bringing local students into our world ; the
letters of rec they get and connections give them newfound possibilities
moving forward.

---

Lillian Martin:

Background/Interests
- Has a natural talent for music
-Junior at Oakland Catholic High School
- Rowing
- Enjoy learning new cultures and languages (Mandarin and Norwegian)
- Likes computer science, technology, cyber security, engineering;
Python and Javascript
- Sketch artist
- Photography
- Filmography and photography
- Mass media


Illah



_______________________________________________
robotics-faculty mailing list
robotics-faculty at lists.andrew.cmu.edu
https://lists.andrew.cmu.edu/mailman/listinfo/robotics-faculty
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20220510/89155df9/attachment.html>

From awd at cs.cmu.edu  Tue May 17 08:33:36 2022
From: awd at cs.cmu.edu (Artur Dubrawski)
Date: Tue, 17 May 2022 08:33:36 -0400
Subject: cameo of software archeology
Message-ID: <CAJvAoys-Gxs48vFNcqhbicMV6vZOe65f7nsEdnr3-OO2WMAhuw@mail.gmail.com>

Team,

I do not think more than a couple of active members of the Auton Lab would
know Jeremy Kubica.
Jeremy was a PhD student and a postdoc here until he left to serve in tech
leadership roles at Google Pittsburgh in June 2006.

As a member of our Lab, he became famous for his incredibly scalable
software code designed to
track asteroids. It helped astrophysicists monitor trajectories of a
couple orders of magnitude
more space rocks than previously possible.

I had no idea that his software from 2005-2006 was still in active use
until a minor complaint received yesterday from one
of the astrophysicists. They reported an issue with recompiling the code
that turned out to be caused by 3rd party changes
to the gmake tool we have been using to build executables of the Auton C
language software in those ancient times.
So, Jeremy's code still works as it always did, and it compiles again
without a glitch after a minor tweak to its make file.

I am tempted to use this story to set the standard of expectation for the
products coming out of our software shop :)

I also would like to congratulate Dr. Kubica for creating a product that
has not aged out of relevance over such a long period of time.

However, the best part of the story is this: Jeremy is back at CMU!
Yes, he has apparently missed a turn on his way back and ended up joining a
different team on campus.
But we will be looking to collaborate and finally bring him fully back to
where he really belongs.

In the meantime, please join me in welcoming Jeremy back and in inviting
him to give a talk at one of our upcoming brainstorming sessions.
It would be great to hear about his past in the Lab, his experience at
Google, and about what he is working on now, or on any other topic
he might want to discuss with us.
(Jeremy: please pick one Monday in the near future, we hold these
brainstorming sessions from 3 to 4pm, in a hybrid mode).

And, maybe we could convince Jeremy to talk about his Auton C libraries at
one of our Coding Coftea meetings?
Jeremy: One of the key points of these meetings is to expose the new
generation of Auton Lab programmers to the wealth of
our C libraries. These meetings take place on Thursdays 12-1pm.

Cheers,
Artur
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20220517/d26233fc/attachment.html>

From jeff4 at andrew.cmu.edu  Wed May 18 09:03:11 2022
From: jeff4 at andrew.cmu.edu (Jeff Schneider)
Date: Wed, 18 May 2022 09:03:11 -0400
Subject: Fwd: Reminder - Thesis Proposal - May 18, 2022 - Ian Char - Quick
 Adaptation via Latent Variables in Model-Based Reinforcement Learning for
 Nuclear Fusion Control
In-Reply-To: <25cf119c-9861-358b-4ebc-940548ea019a@andrew.cmu.edu>
References: <25cf119c-9861-358b-4ebc-940548ea019a@andrew.cmu.edu>
Message-ID: <971158fc-44ac-b068-aba1-204c21379bab@andrew.cmu.edu>

Hi Everyone,

Please come and see Ian's thesis proposal on learning and nuclear 
fusion!  It starts in 28 minutes.

Jeff.



-------- Forwarded Message --------
Subject: 	Reminder - Thesis Proposal - May 18, 2022 - Ian Char - Quick 
Adaptation via Latent Variables in Model-Based Reinforcement Learning 
for Nuclear Fusion Control
Date: 	Tue, 17 May 2022 15:24:58 -0400
From: 	Diane Stidle <stidle at andrew.cmu.edu>
Reply-To: 	stidle at andrew.cmu.edu
To: 	ml-seminar at cs.cmu.edu <ML-SEMINAR at CS.CMU.EDU>, 
riedmiller at google.com, ekolemen at pppl.gov



/*Thesis Proposal*/

Date: May 18, 2022
Time: 9:30 AM (EDT)
Place: GHC 8102 + Remote
Speaker: Ian Char

*Title:?Quick Adaptation via Latent Variables in Model-Based 
Reinforcement Learning for Nuclear Fusion Control*
*
*
Abstract:
Developments in machine learning suggest that learning controllers may 
soon be a viable alternative over the arduous process of developing 
controllers for complicated systems. Control for nuclear fusion is a 
perfect opportunity to employ these algorithms as the dynamics are 
highly non-linear and the physical models that do exist are difficult to 
use for control. In this thesis proposal, I will start by highlighting 
our efforts to learn controllers for fusion with both Bayesian 
optimization and model-based reinforcement learning. As is often the 
case in real world control problems, the performance of these 
controllers are degraded from the fact that the system dynamics of the 
device change from run to run. This is especially the case in nuclear 
fusion as each run of the tokamak may behave differently, which means 
the controller must be able to adapt quickly during operation. To 
address this, I propose modelling the system dynamics using latent 
variables which are observed by the controller and can be quickly 
adjusted for each new run of the device. This quick adaptation is also 
important in cases where the true dynamics are unknown, and I will 
highlight how better epistemic uncertainty modelling can be used to 
improve model-based reinforcement learning. Lastly, I will describe how 
these methods will be leveraged to create powerful controllers for the 
nuclear fusion community.

*Thesis Committee:*
Jeff Schneider (Chair)
Zico Kolter
Ruslan Salakhutdinov
Martin Riedmiller (DeepMind)
Egemen Kolemen (Princeton)

Zoom Meeting link:
https://cmu.zoom.us/j/95486052837?pwd=amNJWlJFUWpYUm9vOGFYVTYvUzduZz09

Link to draft document:
https://drive.google.com/file/d/1oyVdMq6Il2_JXQsDS1J_AkcaIgFP3sEk/view?usp=sharing

-- 
Diane Stidle
Graduate Programs Manager
Machine Learning Department
Carnegie Mellon University
stidle at andrew.cmu.edu


From chiragn at cs.cmu.edu  Thu May 19 17:08:49 2022
From: chiragn at cs.cmu.edu (Chirag Nagpal)
Date: Fri, 20 May 2022 02:38:49 +0530
Subject: Venue for Work in Progress Healthcare Papers
Message-ID: <CAKH1gVoSkC=yck5oOAm-tq_vKnPv2h=P+kr=sbq7qrrmJDN8+A@mail.gmail.com>

Hi All

This might be a decent place to submit smaller or "work in progress"
projects:
https://kdd.org/kdd2022/HealthDay.html

Essentially its a smallish niche workshop to focus on health related topics
held alongside KDD.

-- 

*Chirag Nagpal* PhD Candidate, Auton Lab
School of Computer Science
Carnegie Mellon University
cs.cmu.edu/~chiragn
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20220520/068a74ef/attachment.html>

From jeff4 at andrew.cmu.edu  Tue May 24 09:21:35 2022
From: jeff4 at andrew.cmu.edu (Jeff Schneider)
Date: Tue, 24 May 2022 09:21:35 -0400
Subject: Fwd: RI Ph.D. Thesis Proposal: Viraj Mehta today at 2pm
In-Reply-To: <a28a1b13292a477aaff2b56d06abad2d@cmu.edu>
References: <a28a1b13292a477aaff2b56d06abad2d@cmu.edu>
Message-ID: <f50d4156-25cf-d12d-e089-efc7a799a4dd@andrew.cmu.edu>

Reminder: Come out and hear Viraj's thesis proposal this afternoon on 
Fusion and RL!

Jeff.



-------- Forwarded Message --------
Subject: 	RI Ph.D. Thesis Proposal: Viraj Mehta
Date: 	Wed, 11 May 2022 17:46:04 +0000
From: 	Suzanne Lyons Muth <lyonsmuth at cmu.edu>
To: 	ri-people at lists.andrew.cmu.edu <ri-people at lists.andrew.cmu.edu>



Date:?24?May 2022
Time: 2:00 p.m. (ET)
Location: NSH 4305
Zoom Link: 
https://cmu.zoom.us/j/97423016814?pwd=R2JMRWRhR0w1SUR6S2taQTJQVFhvdz09 
<https://cmu.zoom.us/j/97423016814?pwd=R2JMRWRhR0w1SUR6S2taQTJQVFhvdz09>
Type:?Ph.D.?Thesis Proposal
Who: Viraj Mehta
Title: On Sample-Efficient Reinforcement Learning for Nuclear Fusion

Abstract:

In many practical applications of reinforcement learning (RL), it is 
expensive to observe state transitions from the environment. For 
example, in the problem of plasma control for nuclear fusion, 
determining the next state for a given state-action pair requires 
querying an expensive transition function which can lead to many hours 
of computer simulation or dollars of scientific research. Such expensive 
data collection prohibits application of standard RL algorithms which 
usually require a large number of observations to learn. In this 
proposal, I address the problem of efficiently learning a policy from a 
relatively modest number of observations. The first section leverages 
ideas from Bayesian optimal experimental design to guide the selection 
of state-action queries for efficient learning. The second presents work 
which uses physical prior knowledge about the dynamics to more quickly 
learn an accurate model. Then, I give a brief overview of plasma control 
for nuclear fusion and comment on where I see opportunities for machine 
learning to improve the state of the art for physicists and engineers. I 
present initial work in this direction and give a plan for further 
experiments controlling the rampdown and flattop phases of a plasma shot 
both in simulation and on the DIII-D tokamak. Finally, I discuss my 
timeline for a thesis defense and the projects I would like to 
accomplish along the way.


Thesis Committee Members:
Jeff Schneider, Chair
Deepak Pathak
David Held
Stefano Ermon, Stanford University
Mark D. Boyer, Princeton Plasma Physics Laboratory

A copy of the?thesis?proposal?document is available at:
virajm.com/thesis_proposal.pdf <http://virajm.com/thesis_proposal.pdf>
-------------- next part --------------
_______________________________________________
ri-people mailing list
ri-people at lists.andrew.cmu.edu
https://lists.andrew.cmu.edu/mailman/listinfo/ri-people


From awd at cs.cmu.edu  Sat May 28 12:51:18 2022
From: awd at cs.cmu.edu (Artur Dubrawski)
Date: Sat, 28 May 2022 12:51:18 -0400
Subject: new blog post on Deep VAEs by Ifi
Message-ID: <CAJvAoyvBMXwccsARga+CQ6o-k6cGGG6_dOox60hBTPUd0Eq-aA@mail.gmail.com>

Check this out, it is a very good read:

https://blog.ml.cmu.edu/2022/05/27/deep-attentive-variational-inference/

Cheers,
Artur
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20220528/ed4bf24d/attachment.html>

From predragp at andrew.cmu.edu  Mon May 30 17:41:00 2022
From: predragp at andrew.cmu.edu (Predrag Punosevac)
Date: Mon, 30 May 2022 17:41:00 -0400
Subject: lop2.autonlab.org temp replaced with deimos.autonlab.org
Message-ID: <CAEFFt-GQ-Z35aEOzYYkNgca3P7Eab5CjXDksnF589P+zDG7D_Q@mail.gmail.com>

Dear Autonians,

I hope you have a good Memorial day weekend. Instead of barbequing, I
decided to upgrade our Xen host from Alpine Linux 3.15.4 to 3.16.0.
Unfortunately, things didn't go well and it appears that Xen host is
temporarily broken. For the record the upstream broke things as they rolled
out 3.16.0 without sufficient testing. Obviouly I was dumb enough not to
check things before upgrading.


xen1:/etc/xen/my-guests/auto# xl create lop2.cfg
Parsing config from lop2.cfg
libxl: info: libxl_create.c:121:libxl__domain_build_info_setdefault:
qemu-xen is unavailable, using qemu-xen-traditional instead: No such file
or directory
xc: error: panic: xg_dom_core.c:206: failed to open file
'/usr/lib/xen/boot/ipxe.bin': No such file or directory: Internal error
libxl: error: libxl_dom.c:990:libxl__domain_firmware: failed to load IPXE
/usr/lib/xen/boot/ipxe.bin (-1): No such file or directory
libxl: error: libxl_dom.c:1087:libxl__build_hvm: initializing domain
firmware failed
libxl: error: libxl_create.c:1611:domcreate_rebuild_done: Domain 6:cannot
(re-)build domain: -3
libxl: error: libxl_domain.c:1183:libxl__destroy_domid: Domain
6:Non-existant domain
libxl: error: libxl_domain.c:1137:domain_destroy_callback: Domain 6:Unable
to destroy guest
libxl: error: libxl_domain.c:1064:domain_destroy_cb: Domain 6:Destruction
of domain failed


That affects a handful of lightly used virtual machines but also shell
gateway

lop2.autonlab.org

I quickly changed routing tables so that you can use

deimos.autonlab.org

instead.  That way the number of shell gateways for people who don't have
Auton Lab desktops remains 3 with the caveat that bash is unstable due to
the hardware problems. I will try to replace it in a day or two.

The ssh to demos.autonlab.org will take you to foxconn CPU computing node.
Please don't use that computing node unless you don't have other
alternatives.

Most Kind Regards,
P^2
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20220530/140033de/attachment.html>

From predragp at andrew.cmu.edu  Mon May 30 21:42:20 2022
From: predragp at andrew.cmu.edu (Predrag Punosevac)
Date: Mon, 30 May 2022 21:42:20 -0400
Subject: lop2.autonlab.org temp replaced with deimos.autonlab.org
In-Reply-To: <CAEFFt-GQ-Z35aEOzYYkNgca3P7Eab5CjXDksnF589P+zDG7D_Q@mail.gmail.com>
References: <CAEFFt-GQ-Z35aEOzYYkNgca3P7Eab5CjXDksnF589P+zDG7D_Q@mail.gmail.com>
Message-ID: <CAEFFt-G1-ikARVPrb88BdCseMFRiU1VTjCWgQQiSbQHdaFx9-w@mail.gmail.com>

The issue is fixed! It turns out they were renaming packages and the
firmware we were missing is now part of xen-qemu package.

Predrag

On Mon, May 30, 2022 at 5:41 PM Predrag Punosevac <predragp at andrew.cmu.edu>
wrote:

> Dear Autonians,
>
> I hope you have a good Memorial day weekend. Instead of barbequing, I
> decided to upgrade our Xen host from Alpine Linux 3.15.4 to 3.16.0.
> Unfortunately, things didn't go well and it appears that Xen host is
> temporarily broken. For the record the upstream broke things as they rolled
> out 3.16.0 without sufficient testing. Obviouly I was dumb enough not to
> check things before upgrading.
>
>
> xen1:/etc/xen/my-guests/auto# xl create lop2.cfg
> Parsing config from lop2.cfg
> libxl: info: libxl_create.c:121:libxl__domain_build_info_setdefault:
> qemu-xen is unavailable, using qemu-xen-traditional instead: No such file
> or directory
> xc: error: panic: xg_dom_core.c:206: failed to open file
> '/usr/lib/xen/boot/ipxe.bin': No such file or directory: Internal error
> libxl: error: libxl_dom.c:990:libxl__domain_firmware: failed to load IPXE
> /usr/lib/xen/boot/ipxe.bin (-1): No such file or directory
> libxl: error: libxl_dom.c:1087:libxl__build_hvm: initializing domain
> firmware failed
> libxl: error: libxl_create.c:1611:domcreate_rebuild_done: Domain 6:cannot
> (re-)build domain: -3
> libxl: error: libxl_domain.c:1183:libxl__destroy_domid: Domain
> 6:Non-existant domain
> libxl: error: libxl_domain.c:1137:domain_destroy_callback: Domain 6:Unable
> to destroy guest
> libxl: error: libxl_domain.c:1064:domain_destroy_cb: Domain 6:Destruction
> of domain failed
>
>
> That affects a handful of lightly used virtual machines but also shell
> gateway
>
> lop2.autonlab.org
>
> I quickly changed routing tables so that you can use
>
> deimos.autonlab.org
>
> instead.  That way the number of shell gateways for people who don't have
> Auton Lab desktops remains 3 with the caveat that bash is unstable due to
> the hardware problems. I will try to replace it in a day or two.
>
> The ssh to demos.autonlab.org will take you to foxconn CPU computing
> node. Please don't use that computing node unless you don't have other
> alternatives.
>
> Most Kind Regards,
> P^2
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20220530/0bfdeed5/attachment.html>