去噪迭代自校正：用于可靠LLM推理的结构化验证循环
Denoising Iterative Self-Correction: Structured Verification Loops for Reliable LLM Reasoning
AI总结：
提出DISC方法，通过将验证输出视为噪声测量，在多次验证-判断-校正循环中逐步减少推理错误，并引入二元判断门控机制防止破坏正确答案，在三个基准上优于现有方法。
摘要：
大型语言模型生成流畅但往往不正确的多步推理，而简单的校正方法有降低已正确回答的风险。我们提出去噪迭代自校正（DISC），一种测试时程序，将验证问题输出视为解决方案可能被破坏的噪声测量。利用这些信号，DISC在多次验证-判断-校正传递中逐步减少错误，类似于传统的迭代去噪。二元判断门控通过阻止会破坏已正确回答的重写来控制校正精度，同时验证器和校正器共同修复错误。我们使用两个配对诊断指标评估这一权衡：改进与退化比率（精度）和修复率（召回率）。在三个基准（BIG-Bench Mistake、HotpotQA、GPQA Diamond）和四个模型上，DISC在精度-召回率权衡上优于Chain-of-Verification和Self-Refine，在BIG-Bench Mistake（Sonnet~4.5）上达到81.6%的准确率，每次退化的改进次数比Chain-of-Verification多13倍，比Self-Refine多5倍。在GPQA Diamond上，我们识别出一个能力下限，低于该下限时判断器承认证据中的矛盾但无法将该识别转化为校正。我们进一步表明，跨模型角色分配——将验证和判断分配给与生成器不同的模型——减轻了自我确认偏差。
英文摘要：
Large language models produce fluent but often incorrect multi-step reasoning, and naive correction methods risk degrading already-correct answers. We introduce Denoising Iterative Self-Correction (DISC), a test-time procedure that treats verification question outputs as noisy measurements of where a solution may be corrupted. Using these signals, DISC progressively reduces errors across multiple verify-judge-correct passes, analogous to traditional iterative denoising. A binary judgment gate controls correction precision by blocking rewrites that would damage already-correct answers while the verifier and corrector together repair errors. We evaluate this trade-off using two paired diagnostics: an improvement-to-degradation ratio (precision) and a repair rate (recall). Across three benchmarks (BIG-Bench Mistake, HotpotQA, GPQA Diamond) and four models, DISC dominates Chain-of-Verification and Self-Refine on the precision-recall trade-off, reaching 81.6% accuracy with 13x more improvements per degradation than Chain-of-Verification and 5x more than Self-Refine on BIG-Bench Mistake (Sonnet~4.5). On GPQA Diamond, we identify a capability floor below which judges acknowledge contradictions in evidence but cannot translate that recognition into a correction. We further show that cross-model role allocation -- assigning verification and judgment to a model different from the generator -- mitigates self-confirmation bias.
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。
