Metraipolis › archive › dc_58fb1f2efca5
SWE-Prime: Fewer Trajectories, Better Performance — arXiv:2608.27449
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.27449
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T14:39:27.726Z
- SHA-256 of the archived text
d74b0c4d619bf5d13f8b49671ad8cd0ab75e02128ffc321a9ff44c817a5f43e5- Stable URL of this page
https://metraipolis.com/documento/dc_58fb1f2efca5
How to cite
SWE-Prime: Fewer Trajectories, Better Performance — arXiv:2608.27449 — arXiv (en), https://arxiv.org/abs/2608.27449 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_58fb1f2efca5 (Metraipolis archive dc_58fb1f2efca5, sha256 d74b0c4d619bf5d13f8b49671ad8cd0ab75e02128ffc321a9ff44c817a5f43e5)
BibTeX
@misc{dc58fb1f2efca5,
title = {SWE-Prime: Fewer Trajectories, Better Performance — arXiv:2608.27449},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.27449},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_58fb1f2efca5; sha256 d74b0c4d619bf5d13f8b49671ad8cd0ab75e02128ffc321a9ff44c817a5f43e5},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.27449. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
To improve large language models' ability to resolve real-world software issues, prior work has focused on constructing large-scale agent trajectory datasets and performing supervised fine-tuning (SFT) on successful trajectories. However, task success does not guarantee high-quality supervision: successful trajectories may still contain ineffective, redundant, or risky steps. Directly using such trajectories for SFT can introduce noisy supervision and encourage models to imitate undesirable problem-solving behaviors. Therefore, we propose SWE-Prime, a multi-granularity, two-stage SFT data selection method that progressively filters training data at the trajectory and segment levels. Specifically, the first stage performs trajectory-level screening based on process quality, result quality, and data representativeness, selecting a high-quality and representative subset of successful trajectories. The second stage performs segment-level selection by grouping consecutive steps into semantic segments and assessing each segment based on its contribution to the final solution, learnability, and potential risks. During SFT, all segments remain in the sequence to preserve context, while only selected segments contribute to the loss computation. Experiments on SWE-Bench Pro and SWE-Bench Verified show that training on the 10% trajectory subset selected by SWE-Prime outperforms training on the full resolved dataset, yielding relative performance gains of up to 12.2% and 24.2%, respectively. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-28 in cs.AI). Categorie arXiv: Software Engineering (cs.SE) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2608.27449 — 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 2026-10-17T13:16:53.731Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-28?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).