cs.AINov 17, 2026

RSI-Prime: Training Models to Train Models

Authors: Dewu Zheng, Yang Ye, Hongyu Zhang, Jun Yu, Yanlin Wang

Organizations: Sun Yat-sen University · Mind Lab · Chongqing University · The School of Intelligence Science and Engineering, Harbin Institute of Technology, Shenzhen

Abstract

Enabling large language models (LLMs) to autonomously conduct training and improve target models is an important foundation for recursive self-improvement (RSI). As LLMs expand their role from being trained to conducting training, their model-training capability warrants dedicated study and systematic development. To this end, we construct RSI-Prime, a dataset for improving model-training capability, comprising 350 complete experimental trajectories across seven task categories. To help LLMs learn to make high-quality experimental decisions from training practice, we perform fine-grained decision identification and quality annotation, then apply decision-level loss masking to exclude supervision on low-quality behavior. Experiments with GLM-5.3 on PostTrainBench, a benchmark for autonomous post-training, show that training on RSI-Prime yields a 31.95% relative improvement in overall performance over the original GLM-5.3, highlighting a path toward future RSI systems in which models autonomously sustain their own advancement.

Figures & tables

Appendix figures & tables4 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Training LLM Agents to Automate LLM Post-Training with RSI-Dojo

    Oct 29, 2026Shuzheng Gao, Zhaojian Yu, Penghao Yin +4LLM Post-TrainingLLM Agent Self-Improvement

  2. Language Models Learn to Improve Language Models

    Oct 16, 2026Minhua Lin, Zijun Wang, Zewen Liu +7Language Model Self-ImprovementLarge Language Model-Guided Optimization

  3. RSI-Master: Structuring Experiments to Guide Autonomous Model Improvement

    Sep 28, 2026Yaxin Du, Xiyuan Yang, Zhifan Zhou +10Language Model Post-TrainingSelf-Improving Agents