- ELEPHANT: Measuring and Understanding Social Sycophancy in LLMs
- Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence
- Attention to Non-Adopters
- Cooking Up Creativity: A Cognitively-Inspired Approach for Enhancing LLM Creativity through Structured Representations
- The ML-SUPERB 2.0 challenge
- From Tokens to Thoughts: How LLMs and Humans Trade Compression for Meaning
- Mechanistic evaluation of Transformers and state space models
- Labeling Messages as AI-Generated Does Not Reduce Their Persuasive Effects
- Dynamic Cheatsheet: Test-Time Learning with Adaptive Memory
- CTC-DRO: Robust Optimization for Reducing Language Disparities in Speech Recognition
- Language models cannot reliably distinguish belief from knowledge and fact
Research Experience
Currently a Reynolds Professor in Humanities, Professor of Linguistics, and Professor of Computer Science at Stanford University. Teaching courses include 'Race and Natural Language Processing' (co-taught with Jennifer Eberhardt) and 'From Languages to Information'. Will teach 'The Language of Food' in the future.
Background
Research interests: NLP and its implications for society, and its application to linguistics and other social and cognitive sciences. Past MacArthur Fellow, also works on the language of food.
Miscellany
Personal interest: Research on the language of food