community · 3 papers

Retrieval Augmented Language Models

Retrieval-augmented generation (RAG) and document understanding are the unifying themes, combining large language models with structured retrieval over citation graphs, PDF documents, and knowledge sources. The work spans both methodology—surveying how RAG enhances LLMs and building foundation models over citation graph structures—and evaluation, with benchmarks for assessing diverse PDF document parsing quality.

Papers in this community

  • A Survey on RAG Meeting LLMs: Towards Retrieval-Augmented Large Language Models 2024 DOI ↗
  • OmniDocBench: Benchmarking Diverse PDF Document Parsing with Comprehensive Annotations 2025 DOI ↗
  • LitFM: A Retrieval Augmented Structure-aware Foundation Model For Citation Graphs 2025 DOI ↗

Methods, problems, datasets, and hardware

Neighbouring communities