Metraipolis › archive › dc_4bd6741b67f3

VisualEvolve: A Skill Self-Evolution Framework for Visual Reasoning — arXiv:2610.28386

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.28386
License (as archived)
CC0-1.0
Archived
2026-10-21T07:08:27.645Z
SHA-256 of the archived text
10cf7ff1bd63d7a5e796880ad75a303611cf6c8d85bc331b5f0f79dfa48f684c
Stable URL of this page
https://metraipolis.com/documento/dc_4bd6741b67f3

How to cite

VisualEvolve: A Skill Self-Evolution Framework for Visual Reasoning — arXiv:2610.28386 — arXiv (en), https://arxiv.org/abs/2610.28386 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_4bd6741b67f3 (Metraipolis archive dc_4bd6741b67f3, sha256 10cf7ff1bd63d7a5e796880ad75a303611cf6c8d85bc331b5f0f79dfa48f684c)
BibTeX
@misc{dc4bd6741b67f3,
  title = {VisualEvolve: A Skill Self-Evolution Framework for Visual Reasoning — arXiv:2610.28386},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.28386},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_4bd6741b67f3; sha256 10cf7ff1bd63d7a5e796880ad75a303611cf6c8d85bc331b5f0f79dfa48f684c},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.28386. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Visual tool use enables vision-language models (VLMs) to acquire task-relevant evidence, but its effectiveness can be limited by predefined visual operations and the tool-use strategies learned during training. Representing these operations and strategies as reusable external skills allows them to be refined through self-evolution without updating model parameters. However, skill refinement can introduce performance regressions. Meanwhile, inspecting accumulated task execution trajectories together with skill evolution history may lead to evolution evidence overload. To address these challenges, we propose VisualEvolve, a self-evolving framework that evolves executable visual operations and tool-use strategies for VLMs through a Select--Execute--Evolve--Evaluate cycle. Skill-Library Snapshot Selection uses performance on a fixed validation set disjoint from execution tasks to select an effective historical snapshot for subsequent evolution, limiting the propagation of performance regressions. Progressive Evidence Access (PEA) guides capability-gap diagnosis through a coarse-to-fine hierarchy comprising the Skill-Performance Report, Evolution Log, Trajectory Summaries, and relevant raw trajectory segments. Across four VLM backbones and six evaluation sets from three benchmarks, VisualEvolve improves over the Initial Skill Library in 23 of 24 settings, with a maximum gain of 15.1 percentage points on VisualProbe Hard using Qwen3.5-Plus. Ablation studies support the contributions of Snapshot Selection and PEA, while further experiments show that skills can be improved using experience from different data settings and directly reused across backbones. Moreover, VisualEvolve enables effective skill self-evolution with over an order of magnitude fewer task-solving trajectories than estimated for representative RL-based methods.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.28386 — 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 2026-10-21T07:05:43.647Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).