Vaughn Robinson — accepted AI-conference papers

2025

Aligned LLMs Are Not Aligned Browser Agents

Priyanshu Kumar, Elaine Lau, Saranya Vijayakumar, Tu Trinh, Elaine T Chang, Vaughn Robinson +5

ICLR 2025poster

For safety reasons, large language models (LLMs) are trained to refuse harmful user instructions, such as assisting dangerous activities. We study an open question in this work: does the desired safety refusal, typically enforced in chat contexts, generalize to non-chat and agentic use cases? Unlike…

Cited by 0 Source ↗PDF ↗Scholar ↗

2024

A Careful Examination of Large Language Model Performance on Grade School Arithmetic

Hugh Zhang, Jeff Da, Dean Lee, Vaughn Robinson, Catherine Wu, William Song +9

NeurIPS 2024spotlight

Large language models (LLMs) have achieved impressive success on many benchmarks for mathematical reasoning. However, there is growing concern that some of this performance actually reflects dataset contamination, where data closely resembling benchmark questions leaks into the training data, instea…

Cited by 77 Source ↗PDF ↗Scholar ↗