MixDemo: Enhancing Textual Graph QA with GraphRAG
Discover how MixDemo improves textual graph understanding and question answering using GraphRAG and expert demonstration selection for better AI accuracy.
LLMs vs Human Intelligence in Scientific Discourse
Explore how Large Language Models compare to human intelligence in scientific knowledge and discourse, highlighting their key limitations and challenges.
MDKeyChunker: Efficient Single-Call LLM Enrichment for RAG
Discover MDKeyChunker, a three-stage pipeline enhancing RAG with structure-aware chunking, single-call LLM enrichment, and key-based restructuring for high...
Hierarchical JSON Generation for Scientific Texts Using LLMs
Discover how fine-tuned LLMs generate hierarchical JSON to preserve meaning in scientific sentences, improving text reconstruction and analysis.
Prospective Memory Failures in Large Language Models
Explore how large language models struggle with prospective memory and formatting compliance under complex tasks, impacting AI accuracy and reliability.