Results 71 to 80 of about 186,001 (296)

LLMs Judging LLMs: A Simplex Perspective

Given the challenge of automatically evaluating free-form outputs from large language models (LLMs), an increasingly common solution is to use LLMs themselves as the judging mechanism, without any gold-standard scores. Implicitly, this practice accounts for only sampling variability (aleatoric uncertainty) and ignores uncertainty about judge quality ...
Vossler, Patrick, Xia, Fan, Mai, Yifan, Subbaswamy, Adarsh, Feng, Jean   +4 more
openaire   +2 more sources

Particle-hole Asymmetry of Fractional Quantum Hall States in the Second Landau Level of a Two-dimensional Hole System [PDF]

, 2011
We report the first unambiguous observation of a fractional quantum Hall state in the Landau level of a two-dimensional hole sample at the filling factor $\nu=8/3$.
Csathy, G. A., Kumar, A., Manfra, M. J., Pfeiffer, L. N., Samkharadze, N., West, K. W.   +5 more
core   +3 more sources

Evaluating the Utilities of Foundation Models in Single‐Cell Data Analysis

Advanced Science, EarlyView.
This study delivers the first systematic, task‐level evaluation of single‐cell foundation models across eight core analytical tasks. By benchmarking 10 leading models with the scEval framework, it reveals where foundation models truly add value, where task‐specific methods still dominate, and provides concrete, reproducible guidelines to steer the next
Tianyu Liu, Kexing Li, Yuge Wang, Hongyu Li, Hongyu Zhao   +4 more
wiley   +1 more source

Case Law as Data : Prompt Engineering Strategies for Case Outcome Extraction with Large Language Models in a Zero-Shot Setting

Law, Technology and Humans
This study explores the effectiveness of prompt optimization techniques for legal case outcome extraction using Large Language Models (LLMs). Two state-of-the-art LLMs, LLaMA3 70b and Mixtral 8x7b, are used in a zero-shot data extraction task on a ...
Guillaume Zambrano
doaj   +1 more source

LLM-SQL-Solver: Can LLMs Determine SQL Equivalence?

, 2023
Judging the equivalence between two SQL queries is a fundamental problem with many practical applications in data management and SQL generation (i.e., evaluating the quality of generated SQL queries in text-to-SQL task). While the research community has reasoned about SQL equivalence for decades, it poses considerable difficulties and no complete ...
Zhao, Fuheng, Chen, Jiayue, Lim, Lawrence, Ahmad, Ishtiyaque, Agrawal, Divyakant, Abbadi, Amr El   +5 more
openaire   +2 more sources

High‐Fidelity Synthetic Data Replicates Clinical Prediction Performance in a Million‐Patient Diabetes Cohort

Advanced Science, EarlyView.
This study generates high‐fidelity synthetic longitudinal records for a million‐patient diabetes cohort, successfully replicating clinical predictive performance. However, deeper analysis reveals algorithmic biases and trajectory inconsistencies that escape standard quality metrics. These findings challenge current validation norms, demonstrating why a
Francisco Ortuño, Víctor M. de la Oliva Roque, Javier‐Ignacio Ramirez‐Lopez, David P. Kreil, Joaquín Dopazo, Carlos Loucera   +5 more
wiley   +1 more source

LLM-Supported Manufacturing Mapping Generation [PDF]

Transactions on Graph Data and Knowledge
In large manufacturing companies, such as Bosch, that operate thousands of production lines with each comprising up to dozens of production machines and other equipment, even simple inventory questions such as of location and quantities of a particular ...
Schmidt, Wilma Johanna, Grangel-González, Irlan, Paschke, Adrian, Kharlamov, Evgeny   +3 more
doaj   +1 more source

LLM-AutoDiff: Auto-Differentiate Any LLM Workflow

Large Language Models (LLMs) have reshaped natural language processing, powering applications from multi-hop retrieval and question answering to autonomous agent workflows. Yet, prompt engineering -- the task of crafting textual inputs to effectively direct LLMs -- remains difficult and labor-intensive, particularly for complex pipelines that combine ...
Yin, Li, Wang, Zhangyang
openaire   +2 more sources

Causal Prediction of TP53 Variant Pathogenicity Using a Perturbation‐Informed Protein Language Model

Advanced Science, EarlyView.
A TP53‐specific predictor, CaVepP53, is developed by fine‐tuning ESMC on experimentally validated variants, quantifying pathogenicity via Euclidean distances. It outperforms general‐purpose models and extends to five cancer genes, enabling interpretable variant classification for precision medicine.
Huiying Chen, Yang Zhao, Boqiang Hu, Wuke Wang, Minfang Song, Annabeth Xinyu Zhao, Xiangyang Li, Gefei Wang, Yanfen Wang, Weiyan Zheng, Xinpeng Zhang, Xia Lin, Yanbin Yin, Xingxu Huang, Jinfang Zheng, Tingbo Liang   +15 more
wiley   +1 more source

Large Language Models: А Socio-Philosophical Essay

Галактика медиа: журнал медиа исследований
Neural networks have filled the information space. On the one hand, this indicates the scientific and technological movement of contemporary society (perhaps, AGI is already waiting for us outside the door). On the other hand, in everyday discourse there
Regina V. Penner
doaj   +1 more source