Connectionists: Postdoc at MPI-SWS Saarbrücken in the ChatVerify project for Large Language Models

[Joël Ouaknine]

(see: https://people.mpi-sws.org/~joel/postdoc-CVY.html )

About the project

We invite applications for a postdoctoral research position in Prof. Joël
Ouaknine's research group at the Max Planck Institute for Software Systems
in Saarbrücken, Germany. The successful candidate will work in close
collaboration with academic and industrial partners, delving deep into the
verification of Large Language Models (LLMs) based software programs.

*Focus of the Position:*

   1. Design and implement innovative verification methods to ensure the
   reliability and accuracy of LLM-based software programs, i.e., software
   programs that leverage LLMs to solve various sub-problems.
   2. Actively engage in the design and development of a system that
   generates high-quality data utilising LLMs.

*Project Insight:* Software programs that utilise LLMs to tackle complex
problems often require a strategy to decompose the main problem into
smaller and simpler sub-problems. These sub-problems are separately
addressed using LLMs, and the aggregation of results from these steps
yields the final results. However, employing LLMs in these processes can
lead to uncertainties and errors in the final outcome. Regardless of the
prompt's sophistication, LLMs may generate incorrect responses, which makes
it challenging to ensure the correctness of LLM-based software programs.

Our mission in this project is twofold:

*Prompt-Level Verification:* Establish methods to validate the efficacy of
the prompting processes in getting accurate responses from LLMs.

*Program-Level Verification:* Develop strategies to verify the overall
reliability of the LLM-based software program, specifically its ability to
correctly synthesise intermediate sub-step results even when faced with
potentially uncertain or incorrect results from sub-steps.

Upon establishing a successful verification methodology, the postdoctoral
researcher in this position will further refine this approach, aiding in
the development of a system optimised for high-quality data generation
using LLMs.

While the successful candidate will be hired by, and work at, the Max
Planck Institute for Software Systems in Saarbrücken, collaboration and
frequent visits to research partners are necessary. Such partners include
TU Wien (Vienna, Austria), UCL (London), University of Calabria (Rende,
Cosenza, Italy), as well as industrial collaborators. In addition, the
successful candidate is expected to spend one or more internships in
industry. The project will build on methods and software provided by
industrial partners. We are thus looking for a candidate who is keen and
able to liaise with industry, and who is interested in transformational
research, working on practical problems of industrial relevance.

Your qualifications and responsibilities

*Required:*

   - A PhD degree (or near completion) in algorithmic verification, machine
   learning, databases, or a related field.
   - Strong algorithm-design and coding skills, along with proficiency in
   using popular Machine Learning development frameworks, such as TensorFlow
   and PyTorch.
   - A thorough understanding of the underlying techniques of Large
   Language Models, and experience in fine-tuning or customising such models.
   - Published research articles in top-tier journals or conferences. In
   exceptional cases, industry experience with a solid background in
   industry-based software engineering that has led to highly innovative
   products or results could partially or fully replace the publication
   requirements.
   - Ability and willingness to liaise with industrial partners and to work
   on problems of practical relevance.
   - Interest and ability to perform translational research, applying
   existing theoretical results to practical problems of industrial relevance.
   - You are expected to take on the responsibility of supervising students
   and research assistants.
   - Proficiency in written and spoken English is essential. (Knowledge of
   German is not necessary.)

*Beneficial:*

   - Relevant experience in fields such as Information Extraction from
   Unstructured Text, Automatic Web Data Extraction, Databases, or Fuzzy Logic.
   - Industrial experience, particularly experience in areas such as Big
   Data Engineering and MLOps, coupled with familiarity with cloud services
   such as AWS.
   - Experience with software testing.
   - Experience leading teams or projects, as well as supervising junior
   developers or researchers.
   - A product-oriented mindset.

For informal enquiries, please contact Prof. Joël Ouaknine (joel at mpi-sws.org
).

To apply, please send a cover letter and CV by email to Ms. Annika Meiser (
ameiser at mpi-sws.org).

Applications will be reviewed until a suitable candidate is found. To
ensure full consideration, please submit your application on or before
*22 Oct. 2023*.

We would particularly welcome applications from women and minority
applicants who are currently under-represented within Computer Science.

-- 
*Joël Ouaknine*
Max Planck Institute for Software Systems
Saarland Informatics Campus, Germany
http://mpi-sws.org/~joel/ <http://people.mpi-sws.org/~joel/>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20230907/cd98eb3d/attachment.html>