Enterprise adoption of Large Language Models demands evaluation beyond model accuracy, encompassing cost, latency, compliance, and update adaptability. This paper introduces a Quantitative Evaluation Framework (QEF) to benchmark GPT deployment strategies within Azure OpenAI infrastructure. Five configurations baseline...
Stefan Trajanoski, Aleskandar Karadimce· WSEAS Transactions on Comput...· 0 citations
As large language model-driven AI agents evolve from generative tools into collaborative entities with autonomous capabilities, the human-AI interaction paradigm is shifting from command execution to task delegation. However, empirical evidence on how the behavioral characteristics of AI agents reshape user perception...
Na Liu, Yuxin Zuo· International Journal of Ind...· 0 citations
Large Language Model (LLM)-based generative Information Extraction (IE) over long documents is constrained by finite context windows, and the necessary text-chunking step disrupts semantic continuity, causing subject drift, unresolved coreference, and loss of the global constraints required for fine-grained knowledge g...
Pingfang Tian, Peishan Xie, Qiyuan Li et al.· Scientific Reports· 0 citations
LoRA enables efficient customization of LLMs and is widely used in multi-tenant and multi-task serving. However, emerging model architectures such as MoE significantly increase LoRA memory cost, making existing coupled LoRA serving designs poorly scalable and prone to tail-latency inflation. We present InfiniLoRA, a di...
Hongyu Chen, Letian Ruan, Zilin Xu et al.· Workshop Proceedings of the...· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
Abstract Large Language Models are increasingly used to simulate human opinion dynamics, yet the effect of genuine interaction is often obscured by systematic biases. We develop a Bayesian framework to disentangle and quantify three such biases: (i) A topic bias toward the LLM’s default stance; (ii) an agreement bias f...
Vincent C. Brockers, David A. Ehrlich, Viola Priesemann· Nature Communications· 0 citations
GraphRAG is a multi-component pipeline that also includes query expansion, a second retrieval stage, evidence fusion, and a monotonic RAG fallback, and is therefore evaluated as a complete hybrid system rather than as graph retrieval in isolation, which reflects the complete hybrid system rather than graph augmentation...
Srinivagasam Prabha, B. Collaço, Carina Rosa Malena et al.· Bioengineering· 0 citations
Penicillin allergy is recorded in roughly one in ten hospital patients, yet fewer than one in ten of those records reflects a real, clinically significant allergy. The mismatch matters: patients with a penicillin label receive more second-line and broader-spectrum antibiotics, with measurable downstream costs in infect...
S. Elkhalifa, Jack Devey, Abdulrahaman Al-Asali et al.· Clinical reviews in allergy...· 0 citations
This work introduces holdout asymmetry: the epistemic asymmetry created when a human evaluator can reserve evidence outside a language model's conversational context and later reveal it after the model has committed to an explanatory position. It develops the associated concepts of prediction vacancy and degenerating a...
John F. Ryder· Zenodo (CERN European Organi...· 0 citations
Large language models are increasingly being considered for grant proposal screening, yet their reliability has received less attention than their agreement with human reviewers. This case study evaluates the repeatability and reproducibility of direct textual interpretation by DeepSeek-V4-Flash-0731, run in reasoning...
А. А. Лымарь· Zenodo (CERN European Organi...· 0 citations
Large language models optimise fluently inside a fixed representational space but do not reliably produce genuine discovery; the field has treated this as a tuning problem. We argue it is an architectural one, and we derive the required architecture from first principles rather than inferring it from current systems. T...
Vladimir Mikhailov· Zenodo (CERN European Organi...· 0 citations
Metadata release accompanying the SoftwareX submission. Changed Author affiliations synced with the manuscript, as confirmed by the authors: Fred Torres-Cruz — Doctoral Program in Sustainable Agriculture, UNALM Eduardo Fuentes-Navarro — Faculty of Animal Science, UNALM UNAP department spelled out; corresponding author...
Fred Torres-Cruz, Richar Andre Vilca-Solorzano, Dina Maribel Yana-Yucra et al.· SoftwareX· 0 citations
This collection presents three practical tools developed as part of the Semantic Fidelity Project for evaluating Reality Drift in artificial intelligence systems, including large language models, retrieval systems, and multi-step agent workflows.Reality Drift occurs when a system continues to operate coherently while g...
StudentSim is a research system for evaluating and improving AI tutoring tools before they are tested with learners. It models how individuals may engage with tutors, giving feedback to improve tutor performance across subjects like chess, writing, and mathematics. The post Simulating student learning for better AI tutoring appeared first on Microsoft Research.
MIT News · Artificial Intelligence· news.mit.eduOct 8, 2026
