AI RESEARCH
Why Semantic Entropy Fails: Geometry-Aware and Calibrated Uncertainty for Policy Optimization
arXiv CS.CL
•
Post-