Berkley AI Research

< Back to main page

RL without TD learning
Posted 11-01-2025
What exactly does word2vec learn?
Posted 09-01-2025
Whole-Body Conditioned Egocentric Video Prediction
Posted 07-01-2025
Defending against Prompt Injection with Structured Queries (StruQ) and Preference Optimization (SecAlign)
Posted 04-11-2025
Repurposing Protein Folding Models for Generation with Latent Diffusion
Posted 04-08-2025
Scaling Up Reinforcement Learning for Traffic Smoothing: A 100-AV Highway Deployment
Posted 03-25-2025
Virtual Personas for Language Models via an Anthology of Backstories
Posted 11-12-2024
Linguistic Bias in ChatGPT: Language Models Reinforce Dialect Discrimination
Posted 09-20-2024
How to Evaluate Jailbreak Methods: A Case Study with the StrongREJECT Benchmark
Posted 08-28-2024
Are We Ready for Multi-Image Reasoning? Launching VHs: The Visual Haystacks Benchmark!
Posted 07-20-2024
TinyAgent: Function Calling at the Edge
Posted 05-29-2024
Modeling Extremely Large Images with xT
Posted 03-21-2024
2024 BAIR Graduate Directory
Posted 03-11-2024
The Shift from Models to Compound AI Systems
Posted 02-18-2024
Ghostbuster: Detecting Text Ghostwritten by Large Language Models
Posted 11-14-2023

Discover the groundbreaking research at the Berkeley Artificial Intelligence Research (BAIR) Lab, where UC Berkeley researchers explore computer vision, machine learning, natural language processing, and robotics. With over 50 faculty members and a vibrant community of 300+ graduate students and postdoctoral researchers, BAIR leads the way in advancing AI across various domains.