Metraipolis › archive › dc_fb84312fef2f
SkillCome: Group Contrast Skill Optimization with Dual Memory — arXiv:2609.37128
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.37128
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T16:23:31.028Z
- SHA-256 of the archived text
ce1de483487da702dbe5ff0de73376efc49363365042a54d8b94222bc8786227- Stable URL of this page
https://metraipolis.com/documento/dc_fb84312fef2f
How to cite
SkillCome: Group Contrast Skill Optimization with Dual Memory — arXiv:2609.37128 — arXiv (en), https://arxiv.org/abs/2609.37128 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fb84312fef2f (Metraipolis archive dc_fb84312fef2f, sha256 ce1de483487da702dbe5ff0de73376efc49363365042a54d8b94222bc8786227)
BibTeX
@misc{dcfb84312fef2f,
title = {SkillCome: Group Contrast Skill Optimization with Dual Memory — arXiv:2609.37128},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.37128},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fb84312fef2f; sha256 ce1de483487da702dbe5ff0de73376efc49363365042a54d8b94222bc8786227},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.37128. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Skill evolution improves the capabilities of large language models by analyzing trajectories generated under a given skill and modifying the skill accordingly. Existing approaches typically generate a single trajectory per question. However, this provides insufficient optimization signals since it requires inferring effective skill edits from a solitary path. It is difficult to pinpoint which actions caused the failure in a failed trajectory, or to determine which actions in a successful one should be incorporated into the skill. Furthermore, they rely on a local batch of trajectories for analysis, making the optimization direction susceptible to noisy evidence. To address these, we propose SkillCome, a Skill-evolution method based on group Contrast optimization with dual memory. For each question, SkillCome generates trajectories and performs group contrast analysis to precisely identify key behavioral divergences between successful and failed trajectories, offering reliable optimization signals. The dual memory system further accumulates evidence from historical steps to track patterns shared across different groups, leading to more generalized optimization directions. Together, SkillCome builds a systematic optimization process that transforms experience from observed successful trajectories into reusable skills. Extensive experiments on six benchmarks spanning question answering, reasoning, and agentic tasks demonstrate the effectiveness of our method. SkillCome consistently outperforms baselines across five models of varying families and scales, with gains up to +5.69 points. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.37128 — 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 2026-10-07T15:54:53.591Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).