cs.AINov 7, 2026

No Second Run: Does RSI Stop at Frontier Pre-training?

Authors: Xiaojie Cai, Jiayi Chen, Junwei Chen, Yu Chen, Sikai Cheng, Jinquan Dai, Siyuan Feng, Bin Fu, 

Abstract

Pre-training is where every foundation model begins; bringing recursive self-improvement (RSI) there would let AI improve the process that creates AI. Yet public RSI work has so far targeted only post-training, which starts from an existing base model, runs trials costing GPU-hours, and returns verifiable scores. Pre-training breaks all three conditions at orders-of-magnitude higher cost: a run consumes hundreds of thousands of GPU-hours over months and cannot be repeated; its objective, post-training headroom, cannot be read off any single score; and what must improve is not a model but the whole production system, from data pipelines and training infrastructure to the teams that run them. To our knowledge, we are the first to apply RSI to an entire frontier pre-training run from scratch and to report it in detail. RSI does not stop at frontier pre-training, but it changes form: since no end-to-end loop can close over a single unrepeatable run, we decompose it into smaller loops that can, each repairing a step the full run breaks. The result is 801-30B-A3B-Base, a 30B-A3B MoE trained from scratch on 11.8T tokens in three months by a newly formed team. We release it as a base model; the lightweight SFT reported here (6B tokens, no RL) only probes post-training headroom. As a base model, 801-30B-A3B outperforms Qwen3-30B-A3B and Nemotron-3.5-Lightning-30B-A3B, trailing only the larger Qwen3.5-35B-A3B, and keeps this position under identical SFT; with lightweight SFT alone, it achieves the highest average on four agentic benchmarks against fully post-trained models, including Qwen3.6-35B-A3B. Since post-training RSI presupposes a strong base model, bringing RSI into pre-training acts on the foundation every later stage builds on. We open-source 801-30B-A3B-Base and the lightweight SFT data at https://github.com/GAIR-NLP/801-30B .

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. RSI Arena: Towards Recursive Self-Improvement with Human Feedback

    Oct 24, 2026Yue Huang, Yanbo Wang, Lichi Li +15LLM Agent Self-ImprovementLLM Agent Training