Metraipolis › archive › dc_db38c7ca3cf2
Confidence-Ordering Reversal under Contextual Priors in Neural Decoding — arXiv:2610.08229
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.08229
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T11:19:32.571Z
- SHA-256 of the archived text
91f9af5c283fdc6d36d02d249b318eb3ffa7991ad11570e80acf9facf865d56c- Stable URL of this page
https://metraipolis.com/documento/dc_db38c7ca3cf2
How to cite
Confidence-Ordering Reversal under Contextual Priors in Neural Decoding — arXiv:2610.08229 — arXiv (en), https://arxiv.org/abs/2610.08229 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_db38c7ca3cf2 (Metraipolis archive dc_db38c7ca3cf2, sha256 91f9af5c283fdc6d36d02d249b318eb3ffa7991ad11570e80acf9facf865d56c)
BibTeX
@misc{dcdb38c7ca3cf2,
title = {Confidence-Ordering Reversal under Contextual Priors in Neural Decoding — arXiv:2610.08229},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.08229},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_db38c7ca3cf2; sha256 91f9af5c283fdc6d36d02d249b318eb3ffa7991ad11570e80acf9facf865d56c},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.08229. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Contextual priors improve neural-to-language decoding by reshaping candidate scores. However, confidence is read from the same reshaped scores, so the errors a prior leaves behind can become more confident with no change in accuracy to reveal it. We study how a prior shapes confidence in speech retrieval on MEG-MASC and MOUS using local decoding scores, a contextual prior combined by additive shallow fusion, and the fused top-two margin as confidence. Among initially incorrect predictions, we find a confidence-ordering reversal: a larger margin makes a repair more likely when the correct candidate starts near the top of the local ranking, but less likely when it starts lower. On MEG-MASC, pooled correctness AUROC is 0.87, yet AUROC separating repairs from residual errors falls from 0.70 at initial ranks 2-3 to 0.39 at ranks 21-50. Errors starting beyond rank 20, inside the reversed region, make up 46.6% of all post-fusion errors. We propose a score-level account: a repair must first close the correct candidate's initial deficit, limiting its final margin, whereas a residual error can build a large margin between two incorrect candidates. A causal intervention that changes only the fusion weight moves the reversal to deeper ranks as predicted. Under a word-level LM prior, it keeps moving after accuracy gain peaks, so a weight chosen for accuracy does not settle confidence. Reading local and prior scores separately improves selective decoding: the decoder answers on 74.5% of windows instead of 56.7%, while 92% of output sets still contain the correct candidate. Confidence after contextual fusion should retain the local and contextual evidence behind each prediction, not just the fused scores. Project website: this https URL Code: this https URL Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Artificial Intelligence (cs.AI) ; Information Retrieval (cs.IR); Neurons and Cognition (q-bio.NC). Pagina del lavoro: https://arxiv.org/abs/2610.08229 — 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 2026-10-07T11:15:24.597Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).