Exploration and Exploitation Errors Are Measurable for Language Model Agents Paper • 2604.13151 • Published 3 days ago • 6
Unified Spatio-Temporal Token Scoring for Efficient Video VLMs Paper • 2603.18004 • Published 29 days ago • 13
P1-VL: Bridging Visual Perception and Scientific Reasoning in Physics Olympiads Paper • 2602.09443 • Published Feb 10 • 59