Metraipolis › archive › dc_e7327daa1909

Safe Evolution with Circuit Anchors — arXiv:2608.05158

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.05158
License (as archived)
CC0-1.0
Archived
2026-10-25T21:27:00.654Z
SHA-256 of the archived text
3252d0198b74fe4fa0a9147d4ffe6cf8ea25c94b06386277db7cde592d06c407
Stable URL of this page
https://metraipolis.com/documento/dc_e7327daa1909

How to cite

Safe Evolution with Circuit Anchors — arXiv:2608.05158 — arXiv (en), https://arxiv.org/abs/2608.05158 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e7327daa1909 (Metraipolis archive dc_e7327daa1909, sha256 3252d0198b74fe4fa0a9147d4ffe6cf8ea25c94b06386277db7cde592d06c407)
BibTeX
@misc{dce7327daa1909,
  title = {Safe Evolution with Circuit Anchors — arXiv:2608.05158},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.05158},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e7327daa1909; sha256 3252d0198b74fe4fa0a9147d4ffe6cf8ea25c94b06386277db7cde592d06c407},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.05158. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

In biological evolution, unconstrained mutation can lead to catastrophic outcomes: organisms may evolve enhanced capabilities while losing essential functions for survival. Nature's solution is \textit{developmental constraints}, where core regulatory genes remain anchored while peripheral genes adapt freely. We observe that current self-evolution algorithms for large language models lack analogous constraints. They optimize purely for capability, implicitly assuming safety will be preserved. Our experiments reveal this assumption to be dangerously wrong: models can \textit{misevolve} into powerful yet dangerous entities. Inspired by how Hox genes anchor body structure across $500$ million years of evolution, we propose \textbf{Circuit-Anchored Evolution (CAE)}. Using mechanistic interpretability, we identify a tiny \textit{safety circuit}, comprising less than $2$\% of model features, that causally mediates safety behaviors. We anchor this circuit during evolution, constraining it within a small displacement bound while allowing the remaining features to evolve freely. This mirrors the biological principle of \textit{evolvability with constraint}: preserving what is essential while adapting what is peripheral. Experiments across $3$ model families and two evolution algorithms demonstrate that CAE achieves superior safety preservation with minimal capability loss, substantially outperforming explicit reward-based constraints in both effectiveness and efficiency. Just as developmental constraints prevent biological evolution from producing nonviable organisms, circuit anchoring prevents model evolution from producing capable but dangerous systems.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-07 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL) ; Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE).
Pagina del lavoro: https://arxiv.org/abs/2608.05158 — 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 2026-10-25T21:24:38.167Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-07?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).