CIRAG: Construction-Integration Retrieval and Adaptive Generation for Multi-hop Question Answering

TL;DR

CIRAG introduces a novel multi-hop question answering framework that constructs and integrates evidence iteratively, adaptively expands context, and distills reasoning policies for improved accuracy and efficiency.

Contribution

The paper presents CIRAG, a new model that overcomes greedy expansion limitations and granularity mismatch in multi-hop QA through iterative construction, adaptive generation, and trajectory distillation.

Findings

CIRAG outperforms existing methods on multiple benchmarks.

The model effectively preserves multiple evidence chains.

Trajectory distillation improves reasoning efficiency.

Abstract

Triple-based Iterative Retrieval-Augmented Generation (iRAG) mitigates document-level noise for multi-hop question answering. However, existing methods still face limitations: (i) greedy single-path expansion, which propagates early errors and fails to capture parallel evidence from different reasoning branches, and (ii) granularity-demand mismatch, where a single evidence representation struggles to balance noise control with contextual sufficiency. In this paper, we propose the Construction-Integration Retrieval and Adaptive Generation model, CIRAG. It introduces an Iterative Construction-Integration module that constructs candidate triples and history-conditionally integrates them to distill core triples and generate the next-hop query. This module mitigates the greedy trap by preserving multiple plausible evidence chains. Besides, we propose an Adaptive Cascaded Multi-Granularity Generation module that progressively expands contextual evidence based on the problem requirements, from triples to supporting sentences and full passages. Moreover, we introduce Trajectory Distillation, which distills the teacher model's integration policy into a lightweight student, enabling efficient and reliable long-horizon reasoning. Extensive experiments demonstrate that CIRAG achieves superior performance compared to existing iRAG methods.