Metraipolis › archive › dc_cdc3ea146eb1

InfraRSI: Harnessing Recursive Self-Improvement for Training Infrastructure — arXiv:2611.11558

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.11558
License (as archived)
CC0-1.0
Archived
2026-11-10T18:41:24.084Z
SHA-256 of the archived text
85cfe9b0184b8bdec0031aa6a504e82a0bc1378e134185f27e12833fc155fc5c
Stable URL of this page
https://metraipolis.com/documento/dc_cdc3ea146eb1

How to cite

InfraRSI: Harnessing Recursive Self-Improvement for Training Infrastructure — arXiv:2611.11558 — arXiv (en), https://arxiv.org/abs/2611.11558 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_cdc3ea146eb1 (Metraipolis archive dc_cdc3ea146eb1, sha256 85cfe9b0184b8bdec0031aa6a504e82a0bc1378e134185f27e12833fc155fc5c)
BibTeX
@misc{dccdc3ea146eb1,
  title = {InfraRSI: Harnessing Recursive Self-Improvement for Training Infrastructure — arXiv:2611.11558},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.11558},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_cdc3ea146eb1; sha256 85cfe9b0184b8bdec0031aa6a504e82a0bc1378e134185f27e12833fc155fc5c},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.11558. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

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 $2.75\times$ 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.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2611.11558 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-10T18:37:37.514Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).