Metraipolis › archive › dc_5876ede21074
DIVE: Unlocking Self-Improvement in Frozen Language Models Through Diversity-Driven Skill Evolution — arXiv:2608.12486
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.12486
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-19T04:24:26.525Z
- SHA-256 of the archived text
9b4cec9f65e96c682ec7feb19fdb22de89a65013e30ff6e085d74d55b2fee702- Stable URL of this page
https://metraipolis.com/documento/dc_5876ede21074
How to cite
DIVE: Unlocking Self-Improvement in Frozen Language Models Through Diversity-Driven Skill Evolution — arXiv:2608.12486 — arXiv (en), https://arxiv.org/abs/2608.12486 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5876ede21074 (Metraipolis archive dc_5876ede21074, sha256 9b4cec9f65e96c682ec7feb19fdb22de89a65013e30ff6e085d74d55b2fee702)
BibTeX
@misc{dc5876ede21074,
title = {DIVE: Unlocking Self-Improvement in Frozen Language Models Through Diversity-Driven Skill Evolution — arXiv:2608.12486},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.12486},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5876ede21074; sha256 9b4cec9f65e96c682ec7feb19fdb22de89a65013e30ff6e085d74d55b2fee702},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.12486. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Large language models (LLMs) cannot retain post-deployment experience without parameter updates. We introduce DIVE, a diversity-driven framework that enables frozen LLMs to improve by evolving persistent natural-language skills from task experience and verifier feedback. These skills encode reusable reasoning procedures, verification strategies, common failure modes, and output constraints and are both executed and revised by the same underlying model without access to a teacher model. Since natural-language skill evolution is a stochastic, non-convex search process, optimizing a single skill trajectory can overfit to sampled experience or converge to a suboptimal solution. DIVE mitigates this optimization variance by independently evolving multiple skill populations from bootstrapped experience, adaptively refining them through diverse transformations, and jointly selecting a complementary set of skills. Across six mathematical and logical reasoning tasks and multiple model families, DIVE consistently outperforms existing reasoning methods, prompt-optimization approaches, skill-development frameworks, and memory-based baselines. It achieves rapid self-improvement from accumulated experience, obtaining substantially larger performance gains with fewer rollouts than parameter-based methods such as SFT and GRPO, and prompt optimization with GEPA. Further, the resulting skills transfer across model scales and families, enabling smaller models such as GPT-5-nano to match or outperform larger counterparts, i.e., GPT-5, under conventional prompting. These results establish diversity-driven skill evolution as an effective, interpretable, and parameter-free approach to LLM self-improvement. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-14 in cs.CL). Categorie arXiv: Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2608.12486 — 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 2026-10-19T04:17:42.374Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).