[2026-01-26] Self-Distilled Reasoner: On-Policy Self-Distillation for Large Language Models
OPSD 让同一个模型在两种上下文下分别充当学生和 teacher：学生只看问题并生成自己的轨迹，teacher 额外看到 verified solution，再给同一批 token 密集监督。本文同时核对论文与官…
# On-Policy Distillation
[2026-02-12] On-Policy Context Distillation for Language Models
OPCD 研究怎样把历史经验或优化过的 system prompt 蒸馏进模型参数。它保留 student 的 on-policy rollout，再让 context-conditioned teacher 对同…
[2026-05-09] Training with Harnesses: On-Policy Harness Self-Distillation for Complex Reasoning
OPHSD 把临时 inference harness 变成训练期 teacher context：student 仍直接作答，teacher 则读取 Plan–Solve 或 Draft–Verify 的 ter…
[2026-06-04] ViCuR: Visual Cues as Recoverable Privilege for Multimodal On-Policy Distillation
ViCuR 区分 answer-side privilege 与可从原图恢复的 visual cue，并用 SinkTrack 让一个 sink token 定期读取全体 visual tokens。它适合帮助定义…
[2026-08-05] When Privileged Guidance Misaligns: State-Matched Routing and Contextualized Self-Distillation for Multi-Turn Agents
SMRC-SD 把成功 trajectory 当成 state-indexed resource：只有 student 当前执行状态与 reference prefix 匹配时才启用 dense distillat…
[2026-08-13] Latent On-Policy Self-Distillation
LOPD 检索成功经验并压成 continuous latent tokens，由 latent-conditioned frozen teacher 监督 student on-policy trajectory…
