AI & ML interests

We finetune web-agents with training data (instead of prompting LLMs, and hoping it works!)

Recent Activity

optimass  authored a paper 1 day ago

GitChameleon: Evaluating AI Code Generation Against Python Library Version Incompatibilities

optimass  authored a paper 1 day ago

How to Train Your LLM Web Agent: A Statistical Diagnosis

optimass  authored a paper 1 day ago

LineRetriever: Planning-Aware Observation Reduction for Web Agents

View all activity

Team members 4