Author

Erhan Xu

1 paper indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

#machine learning May 2026

BASIS: Batchwise Advantage Estimation from Single-Rollout Information Sharing for LLM Reasoning

BASIS, a critic-free post-training algorithm designed to address the tradeoff between computational efficiency and sample efficiency in value estimation and policy learning, achieves performance close to multi-rollout GRPO-type baselines and often outperforms single-rollout REINFORCE-type baselines.

Shi-Jin Gong, Erhan Xu, Kai Ye et al. · arXiv.org · 2 citations