Metraipolis › archive › dc_48aaa26325b1
Learning When to Trust Medical Knowledge: Regression-Safe Routing for Clinical Reasoning — arXiv:2610.19663
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.19663
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-15T20:01:53.944Z
- SHA-256 of the archived text
742e3098874d99f745958d2874899ce3a949c6700662f4b41cede729cf4c95c9- Stable URL of this page
https://metraipolis.com/documento/dc_48aaa26325b1
How to cite
Learning When to Trust Medical Knowledge: Regression-Safe Routing for Clinical Reasoning — arXiv:2610.19663 — arXiv (en), https://arxiv.org/abs/2610.19663 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_48aaa26325b1 (Metraipolis archive dc_48aaa26325b1, sha256 742e3098874d99f745958d2874899ce3a949c6700662f4b41cede729cf4c95c9)
BibTeX
@misc{dc48aaa26325b1,
title = {Learning When to Trust Medical Knowledge: Regression-Safe Routing for Clinical Reasoning — arXiv:2610.19663},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.19663},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_48aaa26325b1; sha256 742e3098874d99f745958d2874899ce3a949c6700662f4b41cede729cf4c95c9},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.19663. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Clinical reasoning LLMs need external medical knowledge, but retrieved evidence and reasoning chains can be outdated, poisoned, or hallucinated. Harmful knowledge can cause regression: changing a correct answer into an incorrect one. We construct a medical knowledge-poisoning benchmark covering retrieved evidence and reasoning chains, with seven attack families filtered by a three-judge "subtle-but-wrong" vote and paired clean exams that separate poisoning effects from question difficulty. No fixed knowledge-use strategy is robust across these settings: strong RAG methods trade robustness to poisoned evidence against accuracy when evidence is useful. We formulate knowledge use as per-instance routing and introduce Med-TrustRouter, a lightweight controller that keeps the solver frozen. From leakage-free runtime signals, it estimates each action's expected correctness gain and uncertainty, preserving the base answer unless another action is confidently better. Across seven medical reasoning settings, Med-TrustRouter achieves 78.9 macro and 69.2 worst-case accuracy, matching the strongest competitor on macro accuracy while substantially improving the worst-case floor. Its gains transfer across solvers and evidence judges, unlike a hand-coded router using the same signals. Similar performance across linear, tree, and neural action-value heads suggests that the benefit comes from routing rather than estimator complexity. These results support learning when to trust external knowledge for robust clinical reasoning. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.19663 — 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 2026-10-15T19:35:48.255Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).