Metraipolis › archive › dc_e63b49fd75fd
Artifact-to-Capability: Diagnosing Breaks in Capability Formation of Self-Evolving Agents in a Controlled Environment — arXiv:2611.04952
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.04952
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-04T12:28:40.746Z
- SHA-256 of the archived text
5cdbb041314d49614fe323996b65fb9e861b691be7e22bdb70c7e4fafba62e94- Stable URL of this page
https://metraipolis.com/documento/dc_e63b49fd75fd
How to cite
Artifact-to-Capability: Diagnosing Breaks in Capability Formation of Self-Evolving Agents in a Controlled Environment — arXiv:2611.04952 — arXiv (en), https://arxiv.org/abs/2611.04952 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e63b49fd75fd (Metraipolis archive dc_e63b49fd75fd, sha256 5cdbb041314d49614fe323996b65fb9e861b691be7e22bdb70c7e4fafba62e94)
BibTeX
@misc{dce63b49fd75fd,
title = {Artifact-to-Capability: Diagnosing Breaks in Capability Formation of Self-Evolving Agents in a Controlled Environment — arXiv:2611.04952},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.04952},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e63b49fd75fd; sha256 5cdbb041314d49614fe323996b65fb9e861b691be7e22bdb70c7e4fafba62e94},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04952. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Self-evolving agents write interaction experience into artifacts and reuse them in later tasks. Most work supports such methods with three proxy signals: artifacts accumulate, behavior changes once they are loaded, and metrics rise. We call this signal-based evaluation. What these signals are meant to indicate, however, has not been defined. To fill this gap, we define capability formation from artifacts as five sequentially dependent properties, two of which form a checkpoint for capability formation, and introduce three-scale diagnosis, which judges these properties at the state, game, and lineage levels. In freestyle Gomoku, Reflexion, Voyager, and DGM evolve with the same base model and all show the three signals, yet each exhibits a distinct capability-formation break. Reflexion shows a legalization gain: its pass rate rises only because more of its moves are legal. Voyager shows a compositional override: components that are sound on their own combine to miss all 20 available one-move wins. DGM shows loss-only drift: after passing the checkpoint, later versions only lose passing states. These counterexamples show that signal-based evaluation cannot determine whether a capability has formed. 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/2611.04952 — 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 2026-11-04T12:26:47.185Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).