InfraRSI: Harnessing Recursive Self-Improvement for Training Infrastructure

InfraRSI Team, Yuchuan Tian, Haocheng Mei, Ye Yuan, Chao Xu, Boxun Li, Guohao Dai, Kai Han, Allen Chen, Yunhe Wang, Yu Wang

InfraRSI: Harnessing Recursive Self-Improvement for Training Infrastructure: 32 upvotes on Hugging Face Daily Papers, #7 of 39 papers on 2026-11-10. Day-by-day upvote history.

Upvotes over time

It had 15 upvotes by its first day, 32 upvotes by the next day, 32 upvotes by day 3, 32 upvotes by day 7, and has 32 now. It ranked #21 of 133 papers in its week and #63 of 340 in its month. Its code repository has 47 GitHub stars.

While Recursive Self-Improvement (RSI) accelerates isolated algorithm and kernel synthesis, distributed LLM training depends on tightly coupled launch settings, operator kernels, and communication schedules, where single edits often hit out-of-memory (OOM) boundaries, compiler conflicts, measurement jitter, or silent numerical drift. We present InfraRSI, the agentic harness that turns open-ended human commands into verified training optimizations via a five-stage workflow: init locks editable scope and a fixed task contract over workload, loss, and timing invariants to prevent reward hacking, doctor profiles baseline phases and calibrates cluster noise, run co-optimizes launch knobs and code patches before re-verifying top candidates against the baseline and logging findings to a cross-hardware experience bank, and report and apply audit source paths, track failed-trial costs, and commit promoted changes. Rather than stepping greedily from a single incumbent, InfraRSI initializes from diverse experience-seeded configurations and fits global execution-time and censored peak-memory models across all trials, including OOMs. Using these models, it prices per-setting memory costs, revises miscalibrated LLM priors from live runs, and pairs memory-hungry optimizations with memory savers and verified code patches. Across four single- and multi-node dense and Mixture-of-Experts (MoE) LLM workloads, InfraRSI achieves a speedup of up to 2.75times and lifts mean Model FLOPs Utilization (MFU) from 17.4% to 30.0%, outperforming existing tuners and coding agents while transferring experience across GPUs. As it improves training infrastructure efficiency, InfraRSI forms a two-level Recursive Self-Improvement loop---revising priors online within runs and distilling experience across clusters---to push foundation model scaling.

Paper page on Hugging Face · arXiv

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.