[go: up one dir, main page]
Neil Chowdhury

Hey there! I'm on the founding team of Transluce, a research lab building tools for understanding AI systems. We're hiring!

Previously, I worked on safety at OpenAI (GPT-4o, o1). I graduated from MIT, where I researched AI interpretability & biophysics, and was part of MIT AI Alignment and MIT Video Game Orchestra.

Publications

Surfacing Pathological Behaviors in Language Models

Neil Chowdhury, Sarah Schwettmann, Jacob Steinhardt, Daniel D. Johnson

Automatically Jailbreaking Frontier Language Models with Investigator Agents

Neil Chowdhury, Sarah Schwettmann, Jacob Steinhardt

FrontierScience: Evaluating AI's Ability to Perform Expert-Level Scientific Tasks

Miles Wang, Joy Jiao, Neil Chowdhury, Ethan Chang, Tejal Patwardhan

Investigating truthfulness in a pre-release o3 model

Neil Chowdhury, Daniel D. Johnson, Vincent Huang, Jacob Steinhardt, Sarah Schwettmann

Eliciting Language Model Behaviors with Investigator Agents

Xiang Lisa Li*, Neil Chowdhury*, Daniel D. Johnson, Tatsunori Hashimoto, Percy Liang, Sarah Schwettmann, Jacob Steinhardt

ICML 2025 | arXiv
MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering

Jun Shern Chan*, Neil Chowdhury*, Oliver Jaffe*, James Aung*, Dane Sherburn*, Evan Mays*, Giulio Starace*, Kevin Liu, Leon Maksin, Tejal Patwardhan, Lilian Weng, Aleksander Madry

ICLR 2025 (Oral) | arXiv | GitHub
Introducing SWE-bench Verified

Neil Chowdhury*, James Aung*, Chan Jun Shern*, Oliver Jaffe*, Dane Sherburn*, Giulio Starace*, Evan Mays*, Rachel Dias, Marwan Aljubeh, Mia Glaese, Carlos E. Jimenez, John Yang, Kevin Liu, Aleksander Madry

Nearest Neighbor Normalization Improves Multimodal Retrieval

Neil Chowdhury*, Franklin Wang*, Sumedh Shenoy*, Douwe Kiela, Sarah Schwettmann, Tristan Thrush

EMNLP 2024 | arXiv | GitHub
Building an Early Warning System for LLM-aided Biological Threat Creation

Tejal Patwardhan, Kevin Liu, Todor Markov, Neil Chowdhury, Dillon Leet, Natalie Cone, Caitlin Maltbie, Joost Huizinga, Carroll Wainwright, Shawn (Froggi) Jackson, Steven Adler, Rocco Casagrande, Aleksander Madry

Automatic Discovery of Visual Circuits

Achyuta Rajaram*, Neil Chowdhury*, Antonio Torralba, Jacob Andreas, Sarah Schwettmann

NeurIPS ATTRIB 2023 | arXiv
A Function Interpretation Benchmark for Evaluating Interpretability Methods

Sarah Schwettmann, Tamar R. Shaham, Joanna Materzynska, Neil Chowdhury, Shuang Li, Jacob Andreas, David Bau, Antonio Torralba

NeurIPS 2023 | arXiv | Website
Multimodal Neurons in Pretrained Text-Only Transformers

Sarah Schwettmann*, Neil Chowdhury*, Antonio Torralba

ICCV 2023 Workshop on Closing the Loop Between Vision and Language (Oral) | arXiv | Website
Unique territorial and sub-chromosomal organization revealed in the holocentric moth Bombyx mori

José J. Gil, Leah F. Rosin, Emily Navarette, Neil Chowdhury, Sameer Abraham, Gérard Cornilleau, Elissa P. Lei, Julien Mozziconacci, Leonid A. Mirny, Ines A. Drinnenberg

Diverse Silent Chromatin States Modulate Genome Compartmentalization and Loop Extrusion Barriers

George Spracklin, Nezar Abdennur, Maxim Imakaev, Neil Chowdhury, Sriharsa Pradhan, Leonid A. Mirny, Job Dekker

Projects

Physics is Beautiful
Physics is Beautiful

Interactive physics learning platform

iConverge
iConverge

Web app to help with convergence insufficiency

SecretHitler.io
SecretHitler.io

Web adaptation of the hit social deduction game with 1 million+ games played (contributor)

NetGarage
NetGarage

HackMIT 2022 project illustrating toy neural networks built in 24 hours