新的测试时自我对弈方法在无验证器的情况下扩展了大型语言模型的性能

作者 PulseAugur 编辑部 · [1 个来源] · 2026-11-06 05:00

研究人员开发了一种名为测试时自我对弈（TTSP）的新方法，可以在无需验证器的情况下提高大型语言模型的性能。TTSP 利用非对称自我对弈在部署时扩展性能，即使目标问题没有强大的验证器也适用。在 LiveCodeBench v6 基准测试中，TTSP 在无验证器场景下实现了 35% 的平均绝对增益，并在各种验证器强度下优于其他方法。 AI

影响在没有验证器的部署场景中实现大型语言模型的扩展，有可能提高实际应用性能。

排序理由该条目描述了一篇详细介绍改进大型语言模型新方法的最新研究论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 10 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该条目描述了一篇详细介绍改进大型语言模型新方法的最新研究论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, model release; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Same-day; Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Akira Yoshiyama, Akmal Ashirmatov, Jonas H\"ubotter, Ido Hakimi, Andreas Krause · 2026-11-06 05:00

Test-Time Self-Play: Scaling without a Verifier

arXiv:2611.07367v1 Announce Type: cross Abstract: Test-time scaling methods spend additional compute at deployment to improve a large language model's performance on a hard target problem. Most methods rely on a verifier that scores candidate solutions for correctness, using it t…

本文如何被排名

报道来源 [1]

Test-Time Self-Play: Scaling without a Verifier

相关实体

相关话题