通过节点级一致性检查，低成本精准定位大模型推理错误。
NCV: A Node-Wise Consistency Verification Approach for Low-Cost Structured Error Localization in LLM Reasoning
- 将思维链分解为可验证节点，逐节点做轻量二元校验
- 在公开数据集上F1提升10%~25%，耗能仅传统方法的1/6至1/58
- 适合需要高效可解释推理验证的落地场景
大语言模型多步推理的验证因错误定位不精确且令牌成本高而困难。现有方法或评估整个推理链，导致注意力分散；或依赖昂贵的多次采样。本文提出无需训练的节点级一致性验证（NCV）框架，将验证任务重构为节点层面的轻量级二元一致性检查。通过将思维链分解为相互关联的验证节点，NCV实现错误精准定位，并避免冗余的长文本生成。实验表明，该方法显著提升可解释性与效率，在公开数据集上相较基线提升10%~25%的F1分数，同时使用令牌数仅为基于思维链的验证器的6倍至58倍。
原文摘要 · Abstract (English)
Verifying multi-step reasoning in large language models is difficult due to imprecise error localization and high token costs. Existing methods either assess entire reasoning chains, suffering attention dilution, or rely on expensive multi-sampling. We introduce Node-wise Consistency Verification (NCV), a training-free framework that recasts verification as lightweight binary consistency checks at the node level. By decomposing the chain of thought into interconnected verification nodes, NCV precisely localizes errors and avoids unnecessary long-form generation. Experiments demonstrate that our approach enhances interpretability and efficiency, presenting a scalable solution for reliable LLM reasoning verification. On public datasets, NCV achieves a 10\% to 25\% improvement in F1 scores over baselines while utilizing $6\times$~$58\times$ fewer tokens than traditional methods like CoT-based verifiers.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。