tracked author

Ganqu Cui

Entropy Mechanism equal-contribution author and corresponding author of related RL compositionality / PRIME-RL work. Homepage identifies Tsinghua Ph.D. background, Shanghai AI Lab role, LLM alignment and reinforcement learning research interests, and lists Entropy Mechanism, PRIME, ImplicitPRM and RL compositionality papers.

2 archived notes X: not-found Homepage

Related Notes

按论文归档时间排序，展示该作者在本站已经出现的材料。

归档 Jun 21, 2026

The Entropy Mechanism of Reinforcement Learning for Reasoning Language Models

把 RLVR 训练能力上限关联到 policy entropy 消耗，并分析 advantage update 的熵动力学。

待审阅 2505.22617-entropy-mechanism-rl-reasoning-language-models Training Stability Reasoning RL RL Theory

归档 Mar 20, 2026

From $f(x)$ and $g(x)$ to $f(g(x))$: LLMs Learn New Skills in RL by Composing Old Ones

在受控任务中证明 RL 可组合 base model 已掌握的 atomic skills，形成未见组合能力。

待审阅 2509.25123-rl-compositional-skill-acquisition Reasoning Analysis Reasoning RL Benchmark