IE as Cache: Information Extraction Enhanced Agentic Reasoning

TL;DR

This paper introduces IE-as-Cache, a framework that reuses information extraction as a cognitive cache to improve reasoning accuracy in large language models by maintaining and filtering intermediate information.

Contribution

It proposes a novel approach that repurposes information extraction as a hierarchical cache to enhance multi-step reasoning in language models.

Findings

Significant improvements in reasoning accuracy across diverse benchmarks.

Effective filtering of noise through cache-aware reasoning.

Demonstrated benefits across multiple large language models.

Abstract

Information Extraction aims to distill structured, decision-relevant information from unstructured text, serving as a foundation for downstream understanding and reasoning. However, it is traditionally treated merely as a terminal objective: once extracted, the resulting structure is often consumed in isolation rather than maintained and reused during multi-step inference. Moving beyond this, we propose \textit{IE-as-Cache}, a framework that repurposes IE as a cognitive cache to enhance agentic reasoning. Drawing inspiration from hierarchical computer memory, our approach combines query-driven extraction with cache-aware reasoning to dynamically maintain compact intermediate information and filter noise. Experiments on challenging benchmarks across diverse LLMs demonstrate significant improvements in reasoning accuracy, indicating that IE can be effectively repurposed as a reusable cognitive resource and offering a promising direction for future research on downstream uses of IE.