← Thursday, October 15, 2026AgentsPaper

A 132-run comparison finds limited, unstable gains from agent self-distillation

huggingface.co by Linhao Zhong et al. Editor score 8/10

Across 132 runs and 12,124.4 H800 GPU-hours, ten self-distillation methods produced no consistent gains over GRPO, and strong final scores could hide unstable training. The controlled comparison provides unusually concrete evidence for choosing agent-training methods and budgeting experiments.

What

The study compares ten self-distillation methods with GRPO across four tasks and three model sizes.

Method

Across 132 training runs, it finds gains over GRPO are limited, strong final results can hide unstable training, and extra cost does not consistently improve performance.

Why it matters

Its repeated evaluations and cost measurements give engineers a shared basis for judging agent-training methods and their reliability.

Read on huggingface.co → Code · 1 stars

Headline, summary, brief and diagram are written by a model from the source's own text; the link, byline and numbers come from the source. Read the original before relying on a claim.