4 papers
Don't Make the LLM Read the Graph: Make the Graph Think
Yuqi Sun, Tianqin Meng, George Liu +4
We investigate whether explicit belief graphs improve LLM performance in cooperative multi-agent reasoning. Through 3,000+ controlled trials across four LLM families in the coopera…
Imaginarity witness
Zhiqi Liang, Yufan Lin, Yu Guo +1
Imaginarity has shown to be an important resource in quantum information. The witness theory of quantum resource, such as entanglement witness, coherence witness, and imaginarity w…
Evaluating Control Protocols for Untrusted AI Agents
Jon Kutasov, Chloe Loughridge, Yuqi Sun +4
As AI systems become more capable and widely deployed as agents, ensuring their safe operation becomes critical. AI control offers one approach to mitigating the risk from untruste…
SHADE-Arena: Evaluating Sabotage and Monitoring in LLM Agents
Jonathan Kutasov, Yuqi Sun, Paul Colognese +9
As Large Language Models (LLMs) are increasingly deployed as autonomous agents in complex and long horizon settings, it is critical to evaluate their ability to sabotage users by p…