# Characterizing a Configuration Where Inference-Time PRM-Pruned Fragment Grafting Is Inert: Evidence from Three Reasoning LMs

> Толық мәтін ағылшын тілінде

**Published:** 2026 ж. 2 қазан
**Author:** Khawaja Murad ul Hassan, Mehran Ebrahimi
**Source:** arXiv — cs.AI
**Language:** en-US
**HTML:** https://aiqazaqstan.com/kk/articles/characterizing-a-configuration-where-inference-time-prm-pruned-fragment

Diversity collapse in parallel chain-of-thought has motivated inference-time interventions built on a natural design: when a process reward model (PRM) prunes a chain, its high-PRM prefix is extracted and grafted verbatim as an in-context demonstration into a still-decoding…

> Материал бастапқы дереккөзден алынған: **arXiv — cs.AI**
> [Толық мәтінді arXiv — cs.AI сайтында оқу](https://arxiv.org/abs/2610.00047)

**Тегтер:** #arXiv

---

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