Metraipolis › archive › dc_d796369fd609
Can Coding Agents Sustain Continual Evolution without Forgetting? — arXiv:2611.03219
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.03219
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-03T18:20:23.136Z
- SHA-256 of the archived text
74744672688976b30742f7ca461daf3433a4aa99f0609bf9688231df3f00c102- Stable URL of this page
https://metraipolis.com/documento/dc_d796369fd609
How to cite
Can Coding Agents Sustain Continual Evolution without Forgetting? — arXiv:2611.03219 — arXiv (en), https://arxiv.org/abs/2611.03219 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d796369fd609 (Metraipolis archive dc_d796369fd609, sha256 74744672688976b30742f7ca461daf3433a4aa99f0609bf9688231df3f00c102)
BibTeX
@misc{dcd796369fd609,
title = {Can Coding Agents Sustain Continual Evolution without Forgetting? — arXiv:2611.03219},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.03219},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d796369fd609; sha256 74744672688976b30742f7ca461daf3433a4aa99f0609bf9688231df3f00c102},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.03219. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Recently, harness self-evolution has emerged as an important paradigm for improving agent performance. Prior studies show that evolving the agent harness can effectively enhance agent capabilities without modifying the foundation model's parameters. We find that existing approaches largely study agent evolution on a single task. In practice, coding agents may need to handle multiple task types over time. Whether an agent harness can continuously evolve across multiple tasks without suffering from catastrophic forgetting remains an important open question. In this paper, we investigate this open question through a comprehensive empirical study of continual harness self-evolution. We organize three software engineering task datasets into six different evolution directions and apply two representative harness evolution methods to perform continual evolution along these directions. Our experimental results show that evolution on a subsequent dataset can substantially undermine the performance on previously seen datasets. For example, with DeepSeek-V4-Flash-0731, Meta-Harness improves its performance on SWE-Bench to 94.0% after evolution. However, after further evolution on FeatureBench, its performance on SWE-Bench drops to 32.0% by the fifth round, corresponding to a 66% relative decrease. Finally, based on our experimental findings, we identify several insights, providing directions for the continual self-evolution of agent harnesses. Code is available at this https URL . Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Software Engineering (cs.SE) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2611.03219 — 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 2026-11-03T18:09:53.744Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).