Jacob Andreas

Jacob Andreas is an associate professor in MIT’s Department of Electrical Engineering and Computer Science (EECS), and a member of the Computer Science and Artificial Intelligence Laboratory (CSAIL). His research aims to understand the computational foundations of efficient language learning, and build general-purpose intelligent systems that can communicate effectively with humans and learn from human guidance. He earned a BS from Columbia University and an MPhil from the University of Cambridge, where he studied as a Churchill Scholar. He earned a PhD from the University of California, Berkeley,

Selected Publications

• Jacob, A. P., Shen, Y., Farina, G. & Andreas, J. (2024). The Consensus Game: Language Model Generation via Equilibrium Search. The Proceedings of the International Conference on Learning Representations (ICLR).
• Schwettmann, S., Shaham, T. R., Materzyńska, J., Chowdhury, N., Li, S., Andreas, J., Bau, D. & Torralba, A. (2023). A Function Interpretation Benchmark for Evaluating Interpretability Methods. The Proceedings of the Conference on Neural Information Processing Systems (NeurIPS).
• Hou, B., O’Connor, J., Andreas, J., Chang, S. & Zhang, Y. (2023). PromptBoosting: Black-Box Text Classification with Ten Forward Passes. International Conference on Machine Learning (ICML).

Media

• July 23, 2024: MIT News, MIT researchers advance automated interpretability in AI models
• July 11, 2024: MIT News, Reasoning skills of large language models are often overestimated
• May 14, 2024: MIT News, Using ideas from game theory to improve the reliability of language models