Early-Signal Detector

method · first seen 4 Oct 2026

on-policy self-distillation

Also: OPSD

Training a model by distilling from a teacher that is the same model given privileged information, using on-policy rollouts.

Mentions per week

01419 May11 Aug10 NovPapersCodeDiscussion

PapersCodeDiscussion

Earliest sources

Latest sources

Mentioned together with

group relative policy optimizationreinforcement learning with verifiable rewardscatastrophic forgettingself-evolving agentsknowledge distillationrubric-based rewardmultimodal large language model physics reasoningself-distillation for reasoningcontinual learningmodel merging

In the ledger