Tag: retrieval-augmented generation

A 60-agent LLM Delphi panel matched traditional convergence benchmarks and tracked real publication growth, but standard retrieval left three entire ...

Researchers have built a MetaGPT-based multi-agent AI framework that combines mix-augmented generation with activation-aware weight quantization to deliver fast, accurate ...

A systematic evaluation of 198 Arabic retrieval-augmented generation configurations for Maliki Islamic jurisprudence found a best accuracy of 89.29% using ...

NIEHS researchers have developed Discourse2Draft, a model-agnostic AI application that converts a user query into a structured, citation-anchored draft using ...

A new CARE-Med framework combining dense retrieval, natural language inference verification, and confidence scoring cut unsupported-response rates in medical LLMs ...

Researchers at The Education University of Hong Kong have developed an IoT-enabled multimodal AI framework that identifies food items from ...

A multimodal GraphRAG system tested on electrical-equipment data improved retrieval accuracy and evidence traceability by connecting text, images, structured records, ...

A diabetes question-answering prototype combines a structured medical knowledge graph, guideline retrieval and a large language model while exposing the ...

Researchers describe MyGPT, a fully open-source retrieval-augmented generation platform that runs locally or in secure environments and cites its sources ...

In public-benchmark and retrospective evaluations, HalluciGuard-Eye reduced hallucinations in ophthalmic multimodal AI and maintained diabetic-retinopathy classification performance, without establishing patient-outcome ...