Metraipolis › archive › dc_dfce2ee482ae

Ensemble Methodology for Reference-Free Confidence Estimation in LLM-based Code Summarization — arXiv:2611.09846

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.09846
License (as archived): CC0-1.0
Archived: 2026-11-09T22:34:19.181Z
SHA-256 of the archived text: 2f5e55a64cdf972d247ae901f8d0c7a836765ffe77c4e9aa33be9c7fdeeb422e
Stable URL of this page: https://metraipolis.com/documento/dc_dfce2ee482ae

How to cite

Ensemble Methodology for Reference-Free Confidence Estimation in LLM-based Code Summarization — arXiv:2611.09846 — arXiv (en), https://arxiv.org/abs/2611.09846 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_dfce2ee482ae (Metraipolis archive dc_dfce2ee482ae, sha256 2f5e55a64cdf972d247ae901f8d0c7a836765ffe77c4e9aa33be9c7fdeeb422e)

BibTeX

@misc{dcdfce2ee482ae,
  title = {Ensemble Methodology for Reference-Free Confidence Estimation in LLM-based Code Summarization — arXiv:2611.09846},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.09846},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_dfce2ee482ae; sha256 2f5e55a64cdf972d247ae901f8d0c7a836765ffe77c4e9aa33be9c7fdeeb422e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.09846. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

This paper describes an approach and use case to estimate the correctness of LLM-generated COBOL summaries without access to human-authored references. The core intuition is that inter-model agreement on semantics is informative about truth, and that an LLM-as-judge can provide higher-fidelity pairwise assessments than traditional string- or embedding-based metrics. We operationalise this intuition with (i) controlled summary generation from diverse models; (ii) a LLM evaluator that compares pairs of summaries and emits structured judgments; and (iii) a calibrated aggregation rule that combines supportive evidence while penalising only corroborated contradictions.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Programming Languages (cs.PL).
Pagina del lavoro: https://arxiv.org/abs/2611.09846 — 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 2026-11-09T22:30:47.295Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).