arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2603.03417cs.CRcs.AI
收藏

并行测试时扩展与多序列验证器

Parallel Test-Time Scaling with Multi-Sequence Verifiers

Graduate School of AI, KAIST(人工智能研究生院,韩国科学技术院)

Yegon Kim, Seungyoo Lee, Chaeyun Jang, Hyungi Lee, Juho Lee

AI总结:

提出多序列验证器(MSV),通过条件化候选集预测正确性,改善校准性,提升最佳选择准确率并实现早停策略,在数学推理任务中以不到一半延迟达到相同精度。

摘要:

并行测试时扩展(为单个问题生成多个候选解)是提升大语言模型性能的强大技术。然而,它受到两个关键瓶颈的阻碍:从候选池中准确选择正确的解,以及生成大量完整解带来的高推理延迟。我们认为这两个挑战从根本上与验证器的校准性相关,因为校准良好的验证器能改进答案选择,并支持早停策略以减少延迟。然而,现有的非生成式验证器存在局限性,因为它们孤立地评分每个候选,忽略了候选集之间的丰富上下文信息。为解决这一问题,我们引入了多序列验证器(MSV),这是一种轻量级验证器,它基于完整采样集的条件来预测每个候选的正确性。MSV实现了改进的校准性,这直接增强了最佳N选择性能,并赋能了一种新颖的早停框架。在具有挑战性的数学推理基准测试中,相对于强基线,MSV将最佳64选1的准确率提升了高达6%,并且在早停设置下,以不到一半的延迟达到了与基线相同的准确率。

英文摘要:

Parallel test-time scaling, which generates multiple candidate solutions for a single problem, is a powerful technique for improving large language model performance. However, it is hindered by two key bottlenecks: accurately selecting the correct solution from the candidate pool, and the high inference latency from generating many full solutions. We argue that both challenges are fundamentally linked to verifier calibration, as a well-calibrated verifier improves answer selection and enables early-stopping strategies to reduce latency. However, existing non-generative verifiers are limited as they score each candidate in isolation, overlooking rich contextual information across the set of candidates. To address this, we introduce the Multi-Sequence Verifier (MSV), a lightweight verifier that predicts each candidate's correctness conditioned on the full sampled set. MSV achieves improved calibration, which directly enhances best-of-N selection performance and empowers a novel early-stopping framework. Across challenging mathematical reasoning benchmarks, MSV improves best-of-64 accuracy by up to 6\% relative to strong baselines, and in the early-stopping setting reaches the same accuracy as baselines with less than half the latency.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑