R^3L: Reflect-then-Retry Reinforcement Learning with Language-Guided Exploration, Pivotal Credit, and Positive Amplification Paper • 2601.03715 • Published Jan 7 • 3
SpecBlock: Block-Iterative Speculative Decoding with Dynamic Tree Drafting Paper • 2605.07243 • Published May 8 • 4
TCOD: Exploring Temporal Curriculum in On-Policy Distillation for Multi-turn Autonomous Agents Paper • 2604.24005 • Published Apr 27 • 9