Crimson AI NewsA CrimsonLingua Network service
EN ع
← Back to news
Research paper Hugging Face

Unsupervised Self-Distillation: LLMs Learn from Their Own Majority Votes

AI By Crimson AI Hugging Face Papers 11 August 2026 · 00:00 12 views
Share: X Telegram

A new Hugging Face paper introduces U-OPSD, a method that lets LLMs self-distill without external supervision by using majority-vote pseudo-solutions, improving math reasoning by up to 10.7% over base models.

Unsupervised Self-Distillation: LLMs Learn from Their Own Majority Votes

Key points

On-policy self-distillation has emerged as a promising post-training technique for large language models (LLMs), but most existing approaches depend on external signals such as ground-truth labels, environment feedback, or guidance from larger models. This reliance limits their applicability and moves away from the idea of true self-improvement.

In a new paper, researchers propose Unsupervised On-Policy Self-Distillation (U-OPSD), a method that requires no external supervision at all. Instead, U-OPSD leverages the model's own generations: it samples multiple rollouts, then constructs a pseudo-solution by majority vote under a self-consistency threshold. The model then conditions its distribution on this pseudo-solution and distills itself on the completions where it disagrees, effectively correcting itself precisely where it is confidently wrong.

The method was evaluated across diverse benchmarks, base models, and training settings. On five mathematical reasoning benchmarks (AIME24, AIME25, HMMT25, MATH500, and AMC23), U-OPSD improved over the base model by 8.5% and 10.7% on Qwen3 non-thinking mode at 4B and 8B scales, respectively. It also outperformed supervised OPSD by 3.2% and 2.3% on average in those settings.

In thinking mode, U-OPSD matched OPSD performance, ahead by 0.9% at 4B and level at 8B, while surpassing GRPO by 0.7% and 1.1%, respectively. These results suggest that unsupervised self-distillation can be a viable alternative to supervised methods, potentially reducing the need for labeled data or external feedback in LLM post-training.

The code is available on GitHub, allowing researchers to reproduce and build upon the findings.

SettingScaleImprovement over Base (U-OPSD)vs OPSD (supervised)vs GRPO
Non-thinking4B+8.5%+3.2%
Non-thinking8B+10.7%+2.3%
Thinking4B+0.9%+0.7%
Thinking8Blevel+1.1%
Source
Hugging Face · Hugging Face Papers
Related news
Research paper
Hugging Face 31 Aug 2026

Hugging Face Unveils StepGuard: Step-Level Guardrails for Safer AI Agents

StepGuard, a new step-level guard model from Hugging Face, audits agent actions before execution, reducing attack success rates by...

1
Research paper
Hugging Face 31 Aug 2026

Hugging Face Researchers Unveil ABot-Recon for Stable Long-Horizon 3D Reconstruction

ABot-Recon, a new streaming 3D reconstruction model from Hugging Face, achieves stable long-horizon performance using only local t...

1
Research paper
Hugging Face 31 Aug 2026

ContextPilot: Teaching Agents Proactive Context Management via Fine-Grained RL

Hugging Face researchers introduce ContextPilot, a framework that enhances long-horizon agent reasoning by expanding context-editi...

1