影响力指数

13.5/100

超过 41.6%

全站排名 #37,627

发表论文3 篇

平均评分4.5

年均产出1.5 篇/年

Xiaopeng Li

Assistant Professor@The Chinese University of Hong Kong·中国·OpenReview

Exploration vs Exploitation: Rethinking RLVR through Clipping, Entropy, and Spurious Reward

ICLR 2026Poster

Stepwise Guided Policy Optimization: Coloring your Incorrect Reasoning in GRPO

ICLR 2026Withdrawn

Phase retrieval: Global convergence of gradient descent with optimal sample complexity

ICLR 2025Withdrawn

合作者 (7)

Théodore Fougereux