Metraipolis › archive › dc_05220cd43341
Decision Shifts, Lost Label Functionality, and an Inconclusive Grounding Audit in Correctness-Gated Multi-Teacher Distillation — arXiv:2609.09702
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.09702
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T07:24:29.996Z
- SHA-256 of the archived text
a25616ed6caa95f4081d73175315af8780855ba4ab9bee3ceee0274419f9ec8d- Stable URL of this page
https://metraipolis.com/documento/dc_05220cd43341
How to cite
Decision Shifts, Lost Label Functionality, and an Inconclusive Grounding Audit in Correctness-Gated Multi-Teacher Distillation — arXiv:2609.09702 — arXiv (en), https://arxiv.org/abs/2609.09702 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_05220cd43341 (Metraipolis archive dc_05220cd43341, sha256 a25616ed6caa95f4081d73175315af8780855ba4ab9bee3ceee0274419f9ec8d)
BibTeX
@misc{dc05220cd43341,
title = {Decision Shifts, Lost Label Functionality, and an Inconclusive Grounding Audit in Correctness-Gated Multi-Teacher Distillation — arXiv:2609.09702},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.09702},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_05220cd43341; sha256 a25616ed6caa95f4081d73175315af8780855ba4ab9bee3ceee0274419f9ec8d},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.09702. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Candidate decision correctness and rationale grounding are different objectives. We examine correctness-gated multi-teacher distillation in a fixed experiment. Eight arms share 4,330 sources, a 63.9M-parameter student, 12,990 optimization rows, 406 updates, evidence inputs, and a decoder; seven teacher-based arms use one fixed three-response pool. Three seeds are evaluated on 267 held-out examples. Relative to unfiltered distillation, the correctness-weighted arm differed in accuracy by +0.1660 (95% observed-matrix interval [0.0670, 0.2455]), five-label macro-F1 by +0.1323 ([0.0916, 0.1731]), and task-defined conditional unsafe-action rate by -0.4979 ([-0.5926, -0.3686]). These shifts do not imply uniformly better behavior. Source-label SFT had the highest mean macro-F1 (0.586). The weighted arm had zero Refuted recall in every seed, and two seeds assigned NotEnoughInfo to all 167 claim examples. In an availability-amended audit at one reference seed, weighted and unfiltered outputs had 0/20 versus 1/20 evidence-supported positives and 20/20 versus 19/20 positives containing unsupported material. Samples were non-paired, source overlap was not serialized, and the amendment followed automatic summarization but preceded annotation. The audit therefore cannot estimate a common-source grounding effect and is inconclusive about system-level improvement or harm. Hard filtering already achieved 0.660 accuracy, 0.530 macro-F1, and 0.135 conditional unsafe rate. The implemented weighted arm showed no demonstrated incremental decision benefit over hard filtering. This fixed-matrix failure analysis shows decision redistribution with lost label functionality; the available human audit does not establish a grounding gain. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-10 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.09702 — 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 2026-10-16T07:00:16.208Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-10?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).