**FlowBalance: Verifier-Grounded Self-Improvement from On-Policy Reasoning Exper

**FlowBalance: Verifier-Grounded Self-Improvement from On-Policy Reasoning Experience**

> A reasoning model can improve from its own on-policy experience, but this inner loop is fragile: terminal verifiers provide reliable yet sparse supervision, while dense same-model guidance can reinforce false confidence or overconcentrate learning on a narrow solution mode. We introduce FlowBalance, a verifier-grounded self-improvement method that learns a normalized distribution over complete responses. For each on-...

**来源信息**
- **来源**：HuggingFace Daily Papers（社区热门论文）
- **分类**：行业
- **发布时间**：2026-09-08 08:00（北京时间）
- **原文**：[打开原文](https://huggingface.co/papers/2609.03241)