Belief-Trajectory Energy: Measuring the Path to a Prediction Paper • 2610.05114 • Published 16 days ago • 15
SWE-Bench Pro Verified: A Reliable Benchmark for Software Engineering Agents Paper • 2609.08149 • Published Sep 8 • 27
Spotlight on Token Perception for Multimodal Reinforcement Learning Paper • 2510.09285 • Published Oct 10, 2025 • 37