Job Offer: PhD Causal Machine Learning Applied to NLP and the Study of Large Language Models. Starting date: November 1st, 2023 (flexible) Application deadline: From now until the position is filled Interviews (tentative): Beginning of June and latter if the position is still open Salary: ~2000€ gross/month (social security included) Mission: research oriented (teaching possible but not mandatory) Place of work (no remote): Laboratoire d'Informatique de Grenoble, CNRS, Grenoble, France
Keywords: natural language processing, causal machine learning, interpretability, analysis, robustness, large language models, controllability
Description: Natural language processing (NLP) has undergone a paradigm shift in recent years, owing to the remarkable breakthroughs achieved by large language models (LLMs). Despite being purely "correlation machines" [CorrelationMachine], these models have completely altered the landscape of NLP by demonstrating impressive results in language modeling, translation, and summarization. Nonetheless, the use of LLMs has also surfaced crucial questions regarding their reliability and transparency. As a result, there is now an urgent need to gain a deeper understanding of the mechanisms governing the behavior of LLMs, to interpret their decisions and outcomes in principled and scientifically grounded ways.
A promising direction to carry out such analysis comes from the fields of causal analysis and causal inference [CausalAbstraction]. Examining the causal relationships between the inputs, outputs, and hidden states of LLMs, can help to build scientific theories about the behavior of these complex systems. Furthermore, causal inference methods can help uncover underlying causal mechanisms behind the complex computations of LLMs, giving hope to better interpret their decisions and understand their limitations [Rome].
Thus, the use of causal analysis in the study of LLMs is a promising research direction to gain deeper insights into the workings of these models. As a Ph.D student working on this project, you will be expected to develop a strong understanding of the principles of causal inference and their application to machine learning, see for example the invariant language model framework [InvariantLM]. You will have the opportunity to work on cutting-edge research projects in NLP, contributing to the development of more reliable and interpretable LLMs. It is important to note that the Ph.D. research project should be aligned with your interests and expertise. Therefore, the precise direction of the research can and will be influenced by the personal taste and research goals of the students. It is encouraged that you bring your unique perspective and ideas to the table.
SKILLS Master degree in Natural Language Processing, computer science or data science. Mastering Python programming and deep learning frameworks. Experience in causal inference or working with LLMs Very good communication skills in English, (French not needed).
SCIENTIFIC ENVIRONMENT The thesis will be conducted within the Getalp teams of the LIG laboratory (https://lig-getalp.imag.fr/). The GETALP team has a strong expertise and track record in Natural Language Processing. The recruited person will be welcomed within the team which offer a stimulating, multinational and pleasant working environment. The means to carry out the PhD will be provided both in terms of missions in France and abroad and in terms of equipment. The candidate will have access to the cluster of GPUs of both the LIG. Furthermore, access to the National supercomputer Jean-Zay will enable to run large scale experiments. The Ph.D. position will be co-supervised by Maxime Peyrard and François Portet. Additionally, the Ph.D. student will also be working with external academic collaborators at EPFL and Idiap (e.g., Robert West and Damien Teney)
INSTRUCTIONS FOR APPLYING Applications must contain: CV + letter/message of motivation + master notes + be ready to provide letter(s) of recommendation; and be addressed to Maxime Peyrard (maxime.peyrard@epfl.ch) and François Portet (francois.Portet@imag.fr)
[InvariantLM] Peyrard, Maxime and Ghotra, Sarvjeet and Josifoski, Martin and Agarwal, Vidhan and Patra, Barun and Carignan, Dean and Kiciman, Emre and Tiwary, Saurabh and West, Robert, "Invariant Language Modeling" Conference on Empirical Methods in Natural Language Processing (2022): 5728–5743
[CorrelationMachine] Feder, Amir and Keith, Katherine A. and Manzoor, Emaad and Pryzant, Reid and Sridhar, Dhanya and Wood-Doughty, Zach and Eisenstein, Jacob and Grimmer, Justin and Reichart, Roi and Roberts, Margaret E. and Stewart, Brandon M. and Veitch, Victor and Yang, Diyi, "Causal Inference in Natural Language Processing: Estimation, Prediction, Interpretation and Beyond" Transactions of the Association for Computational Linguistics (2022), 10:1138–1158.
[CausalAbstraction] Geiger, Atticus and Wu, Zhengxuan and Lu, Hanson and Rozner, Josh and Kreiss, Elisa and Icard, Thomas and Goodman, Noah and Potts, Christopher, "Inducing Causal Structure for Interpretable Neural Networks" Proceedings of Machine Learning Research (2022): 7324-7338.
[Rome] Meng, Kevin, et al. "Locating and Editing Factual Associations in GPT." Advances in Neural Information Processing Systems 35 (2022): 17359-17372.