papers

Publications (6)

cs.CY2025

Combining Cost-Constrained Runtime Monitors for AI Safety

Tim Tian Hua, James Baskerville, Henri Lemoine +3

Monitoring AIs at runtime can help us detect and stop harmful actions. In this paper, we study how to efficiently combine multiple runtime monitors into a single monitoring protoco…

cs.AI2026

Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases

Eric Gan, Aryan Bhatt, Buck Shlegeris +2

As AI systems are increasingly used to conduct research autonomously, misaligned systems could introduce subtle flaws that produce misleading results while evading detection. We in…

cs.CL2024

Adaptive Deployment of Untrusted LLMs Reduces Distributed Threats

Jiaxin Wen, Vivek Hebbar, Caleb Larson +9

As large language models (LLMs) become increasingly capable, it is prudent to assess whether safety measures remain effective even if LLMs intentionally try to bypass them. Previou…

cs.CR2026

LinuxArena: A Control Setting for AI Agents in Live Production Software Environments

Tyler Tracy, Ram Potham, Nick Kuhn +31

We introduce LinuxArena, a control setting in which agents operate directly on live, multi-service production environments. LinuxArena contains 20 environments, 1,671 main tasks re…

cs.CR2025

BashArena: A Control Setting for Highly Privileged AI Agents

Adam Kaufman, James Lucassen, Tyler Tracy +2

Future AI agents might run autonomously with elevated privileges. If these agents are misaligned, they might abuse these privileges to cause serious damage. The field of AI control…

cs.LG2025

Ctrl-Z: Controlling AI Agents via Resampling

Aryan Bhatt, Cody Rushing, Adam Kaufman +5

Control evaluations measure whether monitoring and security protocols for AI systems prevent intentionally subversive AI models from causing harm. Our work presents the first contr…