HuggingFace Daily Papers·· 1 天前
TraceForge：用学生模型自身风格生成推理轨迹来做蒸馏
TraceForge: Distilling with Your Own Thinking
AI 导读
TraceForge 通过生成器从上下文和教师答案重新写出学生模型家族风格的推理轨迹，再用判别器判断真实性，绕开对闭源模型隐藏推理的依赖。独立评审给其轨迹的真实性打分平均 0.68，真实轨迹为 0.66，判别器从四个候选中挑出的达 0.82。在相同 5000 道题和教师答案下，HumanEval+ 提升 3.7 分、MBPP+ 提升 7.1 分。
正文 · 原文
Abstract:Using an LLM to generate training data for another is among the most general ways to train one, and it remains common practice. Some frontier teams have closed that route for their own models: what comes back is an answer, or a short summary, without the detailed reasoning trace behind it. That is a little ironic: they were themselves built on distilling human knowledge, and distilling an open-source model is still entirely routine; the restriction applies to their own outputs alone. Recent work recovers the hidden reasoning by replaying encrypted blocks across models. It somehow works, but it breaks the providers' terms of service, so it is not something a product can be built on long-term.
We built the alternative. We take the context and the teacher's answer as input, and generate the reasoning trace ourselves, in the style of the student's own model family. Training is two-stage, supervised and then adversarial: a generator learns to regenerate the family's own reasoning traces from their context and answer, and a critic learns how real a reasoning trace looks. Applied to a teacher's answer, the generator writes one in the family's style. An independent judge rates how likely traces are to be genuine on a scale from 0 to 1: ours average 0.68, real traces 0.66, and traces selected by the critic from four candidates 0.82. These are authenticity scores, not correctness rates.
We test the idea on five benchmarks with everything else held fixed. On general tasks the five training corpora are about the same; on coding, the generated reasoning beats Fable 5's own thinking content. With the same 5,000 questions and the same teacher answers, HumanEval+ improves by 3.7 points and MBPP+ by 7.1 at the final checkpoint, and the MBPP+ lead holds at every checkpoint. We suspect that reasoning written in the student's family style is easier for it to learn from.
| Subjects: | Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.29680 [cs.AI] |
| (or arXiv:2610.29680v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.29680 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yan Wang [view email]
[v1]
Wed, 21 Oct 2026 12:57:29 UTC (1,959 KB)
来源：HuggingFace Daily Papers · arxiv.org