Jacob Andreas

Associate Professor, Department of Electrical Engineering and Computer Science

Computer Science and Artificial Intelligence Laboratory

Natural Language Processing

Machine Learning

External links

Google Scholar

Elements of World Knowledge (EWoK)

Missions

Language

Scaling Inference

Synopsis

Jacob Andreas is interested in language as a communicative and computational tool.

Biography

Jacob Andreas is Associate Professor in MIT's Department of Electrical Engineering and Computer Science, and a member of the Computer Science and Artificial Intelligence Laboratory. 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,

Synopsis

Biography

Publications

(How) Do Language Models Track State?

Learning Linear Attention in Polynomial Time

Elements of World Knowledge (EWOK): A cognition-inspired framework for evaluating basic world knowledge in language models

(How) Do Language Models Track State?

Learning Linear Attention in Polynomial Time

Elements of World Knowledge (EWOK): A cognition-inspired framework for evaluating basic world knowledge in language models

Contextual and combinatorial structure in sperm whale vocalisations

Generative AI in the Era of 'Alternative Facts'

Learning with Language-Guided State Abstractions

Loose LIPS Sink Ships: Asking Questions in Battleship with Language-Informed Program Sampling

Linearity of Relation Decoding in Transformer Language Models

Learning adaptive planning representations with natural language guidance

LILO: Learning Interpretable Libraries by Compressing and Documenting Code

Reasoning or Reciting? Exploring the Capabilities and Limitations of Language Models Through Counterfactual Tasks

What Learning Algorithm is In-Context Learning? Investigations with Linear Models

Joint Modeling of Chest Radiographs and Radiology Reports for Pulmonary Edema Assessment

Compositional Explanations of Neurons

Unnatural Language Processing: Bridging the Gap Between Synthetic and Natural Language Data

Experience grounds language

In the News

Reasoning skills of large language models are often overestimated

Exploring the mysterious alphabet of sperm whales

Natural language boosts LLM performance in coding, planning, and robotics

Reasoning skills of large language models are often overestimated

Exploring the mysterious alphabet of sperm whales

Natural language boosts LLM performance in coding, planning, and robotics

Large language models use a surprisingly simple mechanism to retrieve some stored knowledge

2024 Sloan Research Fellows

Can We Talk to Whales?

3Q: Large Language Models

Researchers discuss frontiers of generative AI

Junior Bose Award for Excellence in Teaching

Solving a machine-learning mystery

Uses of ChatGPT

Rethinking how we communicate with our machines

Teaching machines to reason about what they see

School of Engineering welcomes new faculty

Gallery