Perplexity 与 NVIDIA 提出 JOLT 自蒸馏训练法
摘要
Perplexity 与 NVIDIA AI 提出后训练方法 JOLT，让同一模型同时充当教师与学生，并训练教师依据自身尝试的奖励学习、惩罚偏离学生行为的漂移。在 MATH 基准上，JOLT 以少 6.5 倍的训练 token 达到与 GRPO 相当的准确率；JOLT+ 额外给学生模型结果奖励，在数学、编程、终端任务和工具使用基准上明显优于 GRPO。
摘要由模型根据原文撰写，以原文为准。
正文 · 原文
Sharing new post-training research on self-distillation + RL (JOLT), done together with @NVIDIAAI.
In self-distillation, the same model performs the role of both a teacher and a student. The teacher gets extra information that the student doesn't, and the student learns from feedback provided by the teacher. However, the teacher could still use shortcuts in ways the student has no way to follow.
Our proposed approach, JOLT, trains the teacher as well, to learn from rewards on its own attempts, while penalizing any drift that's far off from what the student model would do. JOLT+ further supplements JOLT by providing the student model outcome-based rewards in addition to the teacher supervision signals.
On the MATH benchmark, we found that JOLT matches GRPO with 6.5x fewer training tokens, while JOLT+ convincingly beats GRPO across math, coding, terminal tasks, and tool use benchmarks.
New research: We developed JOLT, a post-training method where one model acts as both teacher and student, with the teacher trained to give feedback the student can use. On MATH, it matched GRPO's accuracy with 6.5x fewer training tokens. https://www.perplexity.ai/hub/blog/a-good-teacher-meets-the-student-where-they-are在 X 查看被引用的帖子
来源：Aravind Srinivas · x.com
引用原文
AravSrinivas. Sharing new post-training research on self-distillation + RL (JOLT), done together with @NVIDIAAI.[EB/OL]. (2026-10-22)[访问日期]. https://x.com/AravSrinivas/status/2113013527418146913.