# ContractRL: Shielded Group-Relative Policy Optimization for Auditable Tool-Call Repair

> Толық мәтін ағылшын тілінде

**Published:** 2026 ж. 2 қазан
**Author:** Miaobo Hu, Shuhao Hu, Xiaobo Guo, Xin Wang, Bokun Wang, Yina Sa, Daren Zha, Jun Xiao
**Source:** arXiv — cs.AI
**Language:** en-US
**HTML:** https://aiqazaqstan.com/kk/articles/contractrl-shielded-group-relative-policy-optimization-for-auditable-too

Structured tool calls often fail after only a small number of fields violate a schema or an execution contract. Regenerating the complete object enlarges the action surface and makes repeated repair difficult to audit. We introduce ContractRL, a contract-constrained sequential…

> Материал бастапқы дереккөзден алынған: **arXiv — cs.AI**
> [Толық мәтінді arXiv — cs.AI сайтында оқу](https://arxiv.org/abs/2610.00328)

**Тегтер:** #arXiv

---

*AI Qazaqstan — non-commercial informational portal about AI in Kazakhstan. Machine translations and summaries are AI-generated.*