Metraipolis › archive › dc_524daadee644

WeightLoop: Weight-Space Programming with Experimental Feedback — arXiv:2611.07319

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.07319
License (as archived): CC0-1.0
Archived: 2026-11-06T14:50:30.199Z
SHA-256 of the archived text: e7389549d8831fb48b17491d76c6731b6fd7b286b1f0fe3f1c040c18eddbde44
Stable URL of this page: https://metraipolis.com/documento/dc_524daadee644

How to cite

WeightLoop: Weight-Space Programming with Experimental Feedback — arXiv:2611.07319 — arXiv (en), https://arxiv.org/abs/2611.07319 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_524daadee644 (Metraipolis archive dc_524daadee644, sha256 e7389549d8831fb48b17491d76c6731b6fd7b286b1f0fe3f1c040c18eddbde44)

BibTeX

@misc{dc524daadee644,
  title = {WeightLoop: Weight-Space Programming with Experimental Feedback — arXiv:2611.07319},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.07319},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_524daadee644; sha256 e7389549d8831fb48b17491d76c6731b6fd7b286b1f0fe3f1c040c18eddbde44},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.07319. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Weight-space programming (WSP) shows that language models can reason about and rewrite neural network weights; recursive self-improvement (RSI) motivates feedback-driven improvement. Models guiding experiments receive concrete feedback: completed assays reveal prediction errors and inform what to test next. We introduce WeightLoop, a framework that turns experimental feedback into persistent updates to deployed predictors. GPT-6 Astra inspects a task specification and the model implementation, then writes a bounded, reusable parameter-editing program. As new measurements arrive, the program identifies prediction discrepancies, validates candidate edits using observed outcomes, and writes the selected changes back to the original model. The editor remains fixed while the predictor evolves, enabling repeated updates without further LLM calls or gradient retraining while preserving the original inference interface. In offline replay, targeted edits improve mean ranking utility in seven of eight endpoint-budget settings (all eight versus random edits) and mean final discovery across three sequential tasks. WeightLoop connects WSP with an RSI-inspired improvement loop, making experimental evidence a direct driver of model change and subsequent decisions.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2611.07319 — 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 2026-11-06T14:21:42.622Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).