Entropy Ratio Clipping as a Soft Global Constraint for Stable Reinforcement Learning
Paper • 2512.05591 • Published • 17
KlearReasoner
KlearReasoner
-
-
CE-GPPO: Controlling Entropy via Gradient-Preserving Clipping Policy Optimization in Reinforcement Learning
Paper • 2509.20712 • Published • 20 -
Klear-Reasoner: Advancing Reasoning Capability via Gradient-Preserving Clipping Policy Optimization
Paper • 2508.07629 • Published • 43 -
Kwai-Klear/Klear-Reasoner-8B
8B • Updated • 96 • 19 -
Kwai-Klear/KlearReasoner-MathSub-30K
Viewer • Updated • 30k • 112 • 3 -
Kwai-Klear/KlearReasoner-CodeSub-15K
Viewer • Updated • 15k • 36 • 7 -
Kwai-Klear/Klear-Reasoner-8B-SFT
8B • Updated • 29 • 2