LOCKR：一种隐藏状态轨迹引导的规划器，用于检测和修复扩散语言模型中的稳定但错误锁定

中文摘要 AI 辅助

扩散语言模型通过迭代去噪生成文本，在产生最终答案之前会暴露中间轨迹。我们识别出一种反复出现的推理失败，即稳定但错误的锁定，在这种情况中，答案在大量去噪过程尚未完成时便早早地稳定在一个错误值附近。诸如置信度、熵、边际和答案稳定性等表面层面的解码信号，不足以可靠地区分正确的锁定和错误的锁定。我们将选择性推理修复形式化为一个轻量级的测试时规划问题，并提出LOCKR，一种隐藏状态轨迹引导的规划器，它决定何时分配额外的计算资源，扩展一组结构化的针对性修复分支，并使用轨迹感知验证来选择最有希望的延续。在两个扩散语言模型和三个数学推理基准上，隐藏状态轨迹在错误锁定检测和修复选择方面始终优于表面信号和单一隐藏快照。在自然评估分布上，LOCKR在所有五个评估设置中取得了2.21至5.37个百分点的绝对准确率提升，修复率从22%到41%不等。这些结果确立了隐藏扩散轨迹作为选择性测试时推理修复的可操作信号。

英文摘要

Diffusion language models generate text through iterative denoising, exposing intermediate trajectories before final answers are produced. We identify a recurring reasoning failure, stable-but-wrong lock-in, where an answer stabilizes early around an incorrect value while substantial denoising remains. Surface-level decoding signals such as confidence, entropy, margin, and answer stability are insufficient to reliably distinguish correct from erroneous lock-in. We formulate selective reasoning repair as a lightweight test-time planning problem and propose LOCKR, a hidden-state trajectory-guided planner that decides when to allocate additional computation, expands a structured set of targeted repair branches, and selects the most promising continuation using trajectory-aware verification. Across two diffusion language models and three mathematical reasoning benchmarks, hidden-state trajectories consistently outperform surface signals and single hidden snapshots for both wrong-lock-in detection and repair selection. On natural evaluation distributions, LOCKR yields absolute accuracy gains of 2.21--5.37 percentage points across all five evaluated settings, with repair rates ranging from 22% to 41%. These results establish hidden diffusion trajectories as actionable signals for selective test-time reasoning repair.

发表机构

• Rochester Institute of Technology（罗切斯特理工学院）
• NVIDIA Corporation（英伟达公司）

机构由 AI 辅助整理，请以论文原文为准。

补充信息

备注

9 pages, 6 figures, appendix included