Retrieval-Augmented Generation (RAG)

This paper critiques the increasing complexity of LLM agent memory architectures by comparing them against a naive RAG baseline. It reveals that while memory can provide benefits through abstraction, it often imposes a 'tax' by obscuring or losing critical evidence from the raw history.

Chunkr is an ultra-high-performance text-splitting engine written in Rust that optimizes document chunking for LLM and Vector Database pipelines. It utilizes SIMD acceleration and zero-copy memory strategies to deliver massive throughput improvements over existing Python-based libraries.

Canopy is a new framework designed to optimize multimodal RAG by compressing retrieved items into adaptive-granularity regions. It effectively reduces token consumption and improves accuracy by combining hierarchical pruning with critic-guided iterative retrieval.

The authors highlight the lack of a unified definition for 'agentic AI' across legal and technical domains. They propose that a shared taxonomy based on capabilities and system access is essential for effective AI governance.

D-RAC provides a universal pipeline for ingesting complex enterprise documents into RAG systems by normalizing all formats to PDF and using multimodal LLMs for structural conversion. This method significantly reduces the computational cost and latency of document chunking while improving retrieval precision through optimized table and hierarchy handling.

This technical guide critiques the limitations of naive vector-based RAG for complex queries and proposes GraphRAG as a solution. It provides a practical Python implementation using Neo4j to enable multi-hop reasoning through structured knowledge graphs.

This research investigates how different RAG architectures scale in accuracy and cost as document corpora expand from thousands to hundreds of thousands of files. The findings reveal a performance crossover where BM25 becomes the most effective and cost-efficient method at large scales.

This research presents a novel approach to document retrieval that prioritizes the holistic quality of document sets over individual document relevance. By introducing the SetwiseEvalKit benchmark and the Rubric4Setwise optimization method, the authors demonstrate how addressing redundancy and factual conflicts can significantly improve LLM generation performance.