Flexible Entropy Control in RLVR with Gradient-Preserving Perspective Paper • 2602.09782 • Published about 20 hours ago • 2
Metis-SPECS: Decoupling Multimodal Learning via Self-distilled Preference-based Cold Start Paper • 2510.25801 • Published Oct 29, 2025
DEMO: Reframing Dialogue Interaction with Fine-grained Element Modeling Paper • 2412.04905 • Published Dec 6, 2024 • 8