KBQA-R1: Reinforcing Large Language Models for Knowledge Base Question Answering

TL;DR

KBQA-R1 introduces a reinforcement learning framework for knowledge base question answering, enabling LLMs to interactively navigate knowledge graphs with improved accuracy and grounded reasoning.

Contribution

It shifts KBQA from static text imitation to interaction-based learning using reinforcement learning and introduces RRS for better data synthesis.

Findings

Achieves state-of-the-art results on WebQSP, GrailQA, and GraphQuestions.

Effectively grounds LLM reasoning in verifiable execution.

Improves navigation strategies in knowledge graphs.

Abstract

Knowledge Base Question Answering (KBQA) challenges models to bridge the gap between natural language and strict knowledge graph schemas by generating executable logical forms. While Large Language Models (LLMs) have advanced this field, current approaches often struggle with a dichotomy of failure: they either generate hallucinated queries without verifying schema existence or exhibit rigid, template-based reasoning that mimics synthesized traces without true comprehension of the environment. To address these limitations, we present \textbf{KBQA-R1}, a framework that shifts the paradigm from text imitation to interaction optimization via Reinforcement Learning. Treating KBQA as a multi-turn decision process, our model learns to navigate the knowledge base using a list of actions, leveraging Group Relative Policy Optimization (GRPO) to refine its strategies based on concrete execution feedback rather than static supervision. Furthermore, we introduce \textbf{Referenced Rejection Sampling (RRS)}, a data synthesis method that resolves cold-start challenges by strictly aligning reasoning traces with ground-truth action sequences. Extensive experiments on WebQSP, GrailQA, and GraphQuestions demonstrate that KBQA-R1 achieves state-of-the-art performance, effectively grounding LLM reasoning in verifiable execution.