Metraipolis › archive › dc_2fa000b24be0
Do We Need the Largest LLMs for Every Clinical Decision? — arXiv:2610.26937
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.26937
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-21T07:08:43.518Z
- SHA-256 of the archived text
c6a011aac0fb549c188b6d2f044ec969731d59295b971d834918bcd9b4cd25dd- Stable URL of this page
https://metraipolis.com/documento/dc_2fa000b24be0
How to cite
Do We Need the Largest LLMs for Every Clinical Decision? — arXiv:2610.26937 — arXiv (en), https://arxiv.org/abs/2610.26937 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2fa000b24be0 (Metraipolis archive dc_2fa000b24be0, sha256 c6a011aac0fb549c188b6d2f044ec969731d59295b971d834918bcd9b4cd25dd)
BibTeX
@misc{dc2fa000b24be0,
title = {Do We Need the Largest LLMs for Every Clinical Decision? — arXiv:2610.26937},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.26937},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2fa000b24be0; sha256 c6a011aac0fb549c188b6d2f044ec969731d59295b971d834918bcd9b4cd25dd},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.26937. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Large clinical language models often require multi-GPU infrastructure, limiting local deployment. We introduce CERTA, a decision-level statistical certificate that determines when a quantized model may substitute for its native BF16 reference and escalates decisions that cannot be certified to it. Across eleven instruction-tuned models, both general-purpose and domain-adapted, from three families (7-72B), we evaluated 2,500 clinical-verification decisions from de-identified patient contexts (2,484 patients), with patient-disjoint calibration and testing. In the higher-power operational analysis, round-to-nearest 8-bit quantization at BF16 storage served 99.8-100% of held-out decisions across all eleven models; formal finite-sample guarantees were assessed separately using the exact split-sample procedure. In the same operational analysis, the class-conditional 4-bit frontier reduced the amortized, BF16-equivalent deployment cost to 25.8% and 27.3% of the BF16 reference for Qwen2.5-72B and HuatuoGPT-o1-72B, respectively, with measured peak memory of 43.0 and 42.7 GB, below the nominal capacity of a single 48 GB accelerator. Deployment on that specific hardware was not directly tested. Under the exact split-sample certificate, which also accounts for configuration selection, these costs were 26.7% and 27.9%. On external disease named-entity benchmarks against human annotations, the frontier-selected NF4 configurations had a mean absolute F1 change of 0.045 on both datasets, with substantial model-specific variation. CERTA converts compression from a fixed precision choice into a risk-controlled deployment policy for this verification task. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno). Categorie arXiv: Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2610.26937 — 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 2026-10-21T07:05:58.088Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).