8 papers
LLMRouter: Unified Infrastructure for Developing, Evaluating, and Deploying LLM Routers
Tao Feng, Fangxu Yu, Haozhen Zhang +9
No single large language model (LLM) is optimal across all queries and budget constraints, making model routing essential for cost-effective deployment. Existing routers adopt dive…
Interactive Evaluation Requires a Design Science
Keyang Xuan, Peiyang Song, Pan Lu +10
AI evaluation is undergoing a structural change. Large language models (LLMs) are increasingly deployed as systems that act over time through tools, environments, users, and other…
Steer2Adapt: Dynamically Composing Steering Vectors Elicits Efficient Adaptation of LLMs
Pengrui Han, Xueqiang Xu, Keyang Xuan +12
Activation steering has emerged as a promising approach for efficiently adapting large language models (LLMs) to downstream behaviors. However, most existing steering methods rely…
SocialVeil: Probing Social Intelligence of Language Agents under Communication Barriers
Keyang Xuan, Pengda Wang, Chongrui Ye +3
Large language models (LLMs) are increasingly evaluated in interactive environments to test their social intelligence. However, existing benchmarks often assume idealized communica…
TinyScientist: An Interactive, Extensible, and Controllable Framework for Building Research Agents
Haofei Yu, Keyang Xuan, Fenghai Li +6
Automatic research with Large Language Models (LLMs) is rapidly gaining importance, driving the development of increasingly complex workflows involving multi-agent systems, plannin…
Sotopia-RL: Reward Design for Social Intelligence
Haofei Yu, Zhengyang Qi, Yining Zhao +6
Social intelligence has become a critical capability for large language models (LLMs), enabling them to engage effectively in real-world social tasks such as collaboration and nego…