paper
Single and Multi Truth Data Fusion using Large Language Models
paperactiveprovisional
single-and-multi-truth-data-fusion-using-large-language-models-f03332e4·1 events·first seen 16h agoAliases: Single and Multi Truth Data Fusion using Large Language Models
Co-occurring entities
More like this (12)
Multimodal Large Language ModelsExploring Adversarial Robustness and Safety Alignment in Multilingual Multi-Modal Large Language ModelsSame Lesson, Different Story: Cross-Lingual Reconstruction of Cultural Narratives in Large Language ModelsCivil Court Simulation with Large Language Modelslarge language modelsAdaptive Multi-Resolution Procedural Knowledge Compression for Large Language ModelsLarge Language Models (frontier)The Shibboleth Effect: Auditing the Cross-Lingual Distributional Skew of Large Language ModelsSame Evidence, Different Answer: Auditing Order Sensitivity in Multimodal Large Language ModelsThe Riddle Riddle: Testing Flexible Reasoning in Large Language Models and Humansmulti-turn language modelsApparent Psychological Profiles of Large Language Models are Largely a Measurement Artifact
Recent events (1)
LLMs outperform traditional methods on single and multi-truth data fusion tasks
A new arXiv preprint investigates using LLMs for data fusion (truth discovery) over tabular data, covering both single-truth and multi-truth scenarios. The authors evaluate domain-dependent, domain-independent, zero-shot, and one-shot prompting strategies across three benchmark datasets. LLM-based approaches outperform traditional unsupervised methods including DART and LTM on all datasets, with code released publicly.