Postdoc: Reasoning in LLMs

I’m looking for a postdoc to work on reasoning in large language models, with an emphasis on making reasoning more transparent, reliable, and controllable.

Example project areas:

Reasoning evaluation: robust benchmarks/metrics; separating true reasoning from plausible rationalizations.

Steering & control: interventions to improve reasoning quality (prompting, decoding, training-time methods, activation/representation steering).

Faithfulness: detecting unfaithful explanations; increasing explanation–answer alignment.

You can browse my publication list here (if something overlaps with your interests, feel free to reach out):

https://yftah89.github.io/publications

Please reach out with

CV

Publication list

Brief research interest statement (1–2 paragraphs)

Successful candidates typically have publications in top ML/NLP venues (e.g., NeurIPS, ICLR, ICML, ACL, EMNLP, NAACL, AAAI, etc.).