Metraipolis › archive › dc_7c0569220992

Correct but Unselectable: The Hidden Interface Tax in Multi-Candidate Reasoning — arXiv:2610.32235

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.32235
License (as archived)
CC0-1.0
Archived
2026-10-26T20:11:54.890Z
SHA-256 of the archived text
079e9cfb3f9aeceb4c2cc3aff602025ebcc56502b9f4e8cc1d8efad8522aad93
Stable URL of this page
https://metraipolis.com/documento/dc_7c0569220992

How to cite

Correct but Unselectable: The Hidden Interface Tax in Multi-Candidate Reasoning — arXiv:2610.32235 — arXiv (en), https://arxiv.org/abs/2610.32235 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7c0569220992 (Metraipolis archive dc_7c0569220992, sha256 079e9cfb3f9aeceb4c2cc3aff602025ebcc56502b9f4e8cc1d8efad8522aad93)
BibTeX
@misc{dc7c0569220992,
  title = {Correct but Unselectable: The Hidden Interface Tax in Multi-Candidate Reasoning — arXiv:2610.32235},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.32235},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7c0569220992; sha256 079e9cfb3f9aeceb4c2cc3aff602025ebcc56502b9f4e8cc1d8efad8522aad93},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.32235. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Small language models deployed under cost, latency, or privacy constraints cannot simply switch to larger models when reasoning fails. Multi-candidate generate-then-rerank offers a natural test-time scaling route. However, more candidates do not always yield higher accuracy. We identify a previously overlooked bottleneck between generation and selection: a correct answer may appear in the candidate pool yet fail to be selected because it is incomplete, unparseable, or misaligned with the answer mode the fixed selector expects. We term this loss the hidden interface tax and formalize a coverage-conversion decomposition that distinguishes raw oracle coverage from selectable coverage, the subset of problems where a correct candidate can actually be consumed by the selector. We propose Compatibility-Aware Candidate Construction (CACC), an interface layer between the candidate source and the fixed selector that repairs fragments, aligns answer modes, and removes selector-confusing noise without retraining either component. On numeric reasoning (GSM8K, competition_math), CACC raises both oracle coverage and final accuracy. On GPQA Diamond, CACC alone improves final accuracy from 5.6% to 19.2% (+13.6 pp); combined with proposer-side strengthening it reaches 25.8% (+20.2 pp). On MMLU-Pro, CACC exposes a coverage-conversion gap: answer-mode-compatible construction raises oracle coverage, but final accuracy decreases by 0.85 pp under the original selector; adding proposer-side strengthening improves final accuracy.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.32235 — 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 2026-10-26T19:49:03.390Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).