cd/entity/On-Policy Self-distillation· home› entities› On-Policy Self-distillation

grep -l @on-policy self-distillation /news/*.json | wc -l → 1

On-Policy Self-distillation

mentions 1 type Person feed RSS

// recent coverage 1 mentions

04:00

2026-10-12

arxiv.org

large-language-models

Mitigating Capability Degradation in LLM Self-Improvement via Representation Geometry Regularization

A new arXiv paper (2610.12855v1) reports that On-Policy Self-distillation (OPSD) can progressively degrade previously acquired capabilities in large language models, with queries answered correctly at…

// co-occurs with top 3 entities

arXiv 1 Representation Geometry Regularization 1 supermachinelearner/RGR 1

// topics top 4 topics

large language models 1 machine learning 1 ai research 1 artificial intelligence 1