2 papers
cs.LG2026
Triage-PiEvo: Evidence Admission for Principle-Evolving Scientific Agents
Peize Li, Yingming Pu, Xingguo Xu +3
Large language model (LLM)-based AI Scientists advance scientific discovery, yet they suffer from inefficiencies due to noisy observations. Existing principle-evolving agents act o…
cs.RO2026
CriticHack: Evaluating Visual Rewards Under Robot Policy Optimization
Jiaxuan Luo, Xingguo Xu, Shanshan Wang +2
Learned visual reward models are increasingly used to optimize robot policies, yet a reward model can score an execution that acts on the wrong object as highly as one that complet…