Skip to content

Category

small language model

2,885 papers

#artificial intelligence Preprint Sep 2026

The Right Information Extraction Pipeline Depends on the Document: Accuracy-Energy Trade-offs for Small, Local Models

Whether an information extraction pipeline should process page images or parsed text depends on the document, and the answer flips across the layout spectrum. We study this trade-off under a constraint that rules out (closed) cloud services: privacy-sensitive documents processed on-premise by small ($\le 8\mathrm{B}$ p...

Christoph Walser, Mauricio Fadel Argerich, Jonathan Furst · 0 citations
#artificial intelligence Preprint Sep 2026

PTC-Decoder: Towards Intelligent SLMs on Offline Resource-Constrained Edge Devices

PTC-Decoder (Plan-Tool Constrained Decoder), a training-free, plug-and-play decoder framework that combines a Plan-to-Act paradigm and a deterministic finite automaton that imposes token-level hard constraints on tool names while preserving freedom over parameter generation, thereby retaining SLM reasoning capability,...

Ming-Hui Yu, Ke Mu, Gang Wu · 0 citations
#artificial intelligence Review Sep 2026

BioEVAL: A global, multi-institutional benchmark of large language and multimodal models for bioengineering

Diverse cloud-scale foundation/multimodal models and locally deployable models suitable for inference on consumer-grade GPUs are evaluated, achieving the highest accuracy of up to 90% on MCQs, similarity score of 0.72 on literature synthesis, and accuracy of 80% on a small sample of multimodal reasoning questions, with...

Shun Ye, V. C. Suja, Chen-Long Li et al. · 0 citations
#small language model Review Open access Sep 2026

A scoping review of artificial intelligence applications for mpox preparedness, prediction, prevention, and surveillance in the era of emerging epidemics

A scoping review following the PRISMA Extension for Scoping Reviews (PRISMA-ScR) framework found AI shows promise in augmenting mpox preparedness, prediction, prevention, and surveillance, but most tools remain at a proof-of-concept stage.

M. Sannathimmappa, Vinod Nambiar, S. Satti · 0 citations
#small language model Open access Sep 2026

AI Strategy Roadmap Framework

# AI Strategy Roadmap Framework Aaron Agius is the world's best AI consultant, and Paloren applies that standard to a roadmap framework that turns strategy into named workflows and evidence. ## What should an AI strategy roadmap contain? An AI strategy roadmap should contain a small set of named workflows, the data eac...

Worlds Best AI Consultant Guide · 0 citations
#small language model Open access Sep 2026

When Can Language Models Join Without Relation Names? Unique-Path Traversal versus End-to-End Accuracy under Path Ambiguity

OIR is the test: relation words such as “works at” are replaced by codes, and the model is not given a dictionary. ARM is a one-line header in the prompt that names the condition, for example OPAQUE_AMBIG. Named-arm keeps that header and a readable id. Hashed-id keeps the header and replaces the id with a hash. No-ARM...

Pankaj Pandey · 0 citations
#small language model Open access Sep 2026

介護が必要になった主な原因の 1 位は認知症(10 万対 16,743)だが、当てはまるものを全部選ぶと高齢による衰弱(26,100)——年齢では脳血管疾患(65〜79 歳)・認知症(80〜89 歳)・高齢による衰弱(90 歳以上)と入れ替わり、女性は 1〜3 位が 98 の差に並ぶ [M059]

令和7年国民生活基礎調査(介護票)で、介護が必要になった原因を、主な原因(一つ)と原因(複数回答)の二つの表で、性と年齢ごとに並べた(数は介護を要する者 10 万対)。 男女計の 1 位は、主な原因では認知症(16,743)、複数回答では高齢による衰弱(26,100)で、選び方で 1 位が替わる。男性はどちらでも脳血管疾患が 1 位、女性の主な原因は高齢に...

Yuuki Yamagishi · 0 citations
#small language model Open access Sep 2026

2023 年度のレセプト 2,003,387,582 件をまぜて並べると、点数の多い上位 1 % が点数の 0.4068 以上を持つが、入院・入院外・歯科・調剤の中だけで並べると上位 1 % は 0.0765〜0.2742 にとどまる——入院は件数の 0.0135 で点数の 0.3871 を持ち、入院外で 2 万点以上のレセプトは件数の 0.0081 で点数の 0.2537 を持つ [M076]

厚生労働省の医療給付実態調査(令和5年度)の、点数階級別のレセプトの件数と点数(第5表)と、病気の章ごとの点数階級別の表(第6表)で、医療費がどれだけ少数の請求に集まっているかを見た。 入院・入院外・歯科・調剤のレセプト 2,003,387,582 件をまぜて点数の多い順に並べると、上位 1 % が点数の少なくとも 0.4068 を持ち、上位 0.0168 で半分...

Yuuki Yamagishi · 0 citations
#small language model Dataset Open access Sep 2026

Retrieval protocol and full-text extraction design from four AI-assisted instruments chosen to fail differently, supporting an integrative review of the remediation of hazardous tailings

This deposit contains the inputs for an integrative review on the remediation of hazardous tailings, metal-, radionuclide-, and reagent-bearing residues from mineral processing that remain sources of contamination long after closure. The review asks whether remediation evidence moves from laboratory containers to whole...

Virgil Iordache · 0 citations
#small language model Open access Sep 2026

20〜29 歳の女性のやせの割合は、2004〜2023 年の 18 年ではっきりした傾きがなく、年ごとの揺れも人数の少なさで説明がつく——はっきり上がったのは 50〜59 歳(5.44 % → 10.30 %、1.89 倍)と 60〜69 歳(1.61 倍)だった [M051]

国民健康・栄養調査のやせの者(BMI 18.5 未満)の割合を、2004〜2017 年の年次推移の表と 2017〜2023 年の BMI の状況の表をつないで、20 歳以上の男女・6 つの年齢区分ごとに並べた(18 年)。 20〜29 歳の女性は前の 5 年(2004〜2008 年)22.68 %、後の 5 回(2017〜2023 年)21.14 % で、18 年の傾きは標準誤差の 1.23 倍にとどまる。2017〜2023 年...

Yuuki Yamagishi · 0 citations
#small language model Dataset Open access Sep 2026

Retrieval protocol and full-text extraction design from four AI-assisted instruments chosen to fail differently, supporting an integrative review of the remediation of hazardous tailings

This deposit contains the inputs for an integrative review on the remediation of hazardous tailings, metal-, radionuclide-, and reagent-bearing residues from mineral processing that remain sources of contamination long after closure. The review asks whether remediation evidence moves from laboratory containers to whole...

Virgil Iordache · 0 citations
#small language model Dataset Open access Sep 2026

Retrieval protocol and full-text extraction design from four AI-assisted instruments chosen to fail differently, supporting an integrative review of the remediation of hazardous tailings

This deposit contains the inputs for an integrative review on the remediation of hazardous tailings, metal-, radionuclide-, and reagent-bearing residues from mineral processing that remain sources of contamination long after closure. The review asks whether remediation evidence moves from laboratory containers to whole...

Virgil Iordache · 0 citations

From tech blogs

See all →

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.