Metraipolis › archive › dc_2d9bf33abd77
Enhancing Small Language Models for Power Outage Report Generation via Minimum Risk Training — arXiv:2609.27197
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.27197
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T09:11:08.583Z
- SHA-256 of the archived text
ec1ad28c64cb9b63e2706065d7fce6cf6f84ec58364278dc42af5a06cce646d4- Stable URL of this page
https://metraipolis.com/documento/dc_2d9bf33abd77
How to cite
Enhancing Small Language Models for Power Outage Report Generation via Minimum Risk Training — arXiv:2609.27197 — arXiv (en), https://arxiv.org/abs/2609.27197 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2d9bf33abd77 (Metraipolis archive dc_2d9bf33abd77, sha256 ec1ad28c64cb9b63e2706065d7fce6cf6f84ec58364278dc42af5a06cce646d4)
BibTeX
@misc{dc2d9bf33abd77,
title = {Enhancing Small Language Models for Power Outage Report Generation via Minimum Risk Training — arXiv:2609.27197},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.27197},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2d9bf33abd77; sha256 ec1ad28c64cb9b63e2706065d7fce6cf6f84ec58364278dc42af5a06cce646d4},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.27197. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Minimum Risk Training (MRT) enables neural machine translation models to directly optimize sequence-level evaluation metrics instead of relying only on token- level maximum-likelihood objectives Shen et al. [2016]. Although introduced a decade ago, recent work shows renewed potential for risk-based optimization in modern language models Yang et al. [2024], Jinnai et al. [2025]. We apply MRT to power outage report generation for the Outage Data Initiative Nationwide (ODIN), transforming heterogeneous reports into standardized XML compliant with CIM IEC 61968-3. Our MRT approach improves Qwen2.5-7B-Instruct overall accuracy from 16.20% to 68.95%, demonstrating the effectiveness of sequence- level optimization for domain-specific structured generation Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-24 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.27197 — 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 2026-10-08T08:13:26.475Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).