Pruning Minimal Reasoning Graphs for Efficient Retrieval-Augmented Generation
This addresses the problem of high computational cost and latency in RAG systems for knowledge-intensive LLM tasks, offering a practical solution for long-running sessions and evolving corpora, though it appears incremental as it builds on existing RAG and graph methods.
The paper tackles the inefficiency of retrieval-augmented generation (RAG) systems by introducing AutoPrunedRetriever, a graph-style system that persists and incrementally extends minimal reasoning subgraphs, reducing redundant retrieval and reasoning. It achieves state-of-the-art accuracy, improving over HippoRAG2 by 9-11 points on benchmarks, while using up to two orders of magnitude fewer tokens than baselines.
Retrieval-augmented generation (RAG) is now standard for knowledge-intensive LLM tasks, but most systems still treat every query as fresh, repeatedly re-retrieving long passages and re-reasoning from scratch, inflating tokens, latency, and cost. We present AutoPrunedRetriever, a graph-style RAG system that persists the minimal reasoning subgraph built for earlier questions and incrementally extends it for later ones. AutoPrunedRetriever stores entities and relations in a compact, ID-indexed codebook and represents questions, facts, and answers as edge sequences, enabling retrieval and prompting over symbolic structure instead of raw text. To keep the graph compact, we apply a two-layer consolidation policy (fast ANN/KNN alias detection plus selective $k$-means once a memory threshold is reached) and prune low-value structure, while prompts retain only overlap representatives and genuinely new evidence. We instantiate two front ends: AutoPrunedRetriever-REBEL, which uses REBEL as a triplet parser, and AutoPrunedRetriever-llm, which swaps in an LLM extractor. On GraphRAG-Benchmark (Medical and Novel), both variants achieve state-of-the-art complex reasoning accuracy, improving over HippoRAG2 by roughly 9--11 points, and remain competitive on contextual summarize and generation. On our harder STEM and TV benchmarks, AutoPrunedRetriever again ranks first, while using up to two orders of magnitude fewer tokens than graph-heavy baselines, making it a practical substrate for long-running sessions, evolving corpora, and multi-agent pipelines.