Research map / Large language models / Retrieval and knowledge
KBQA-R1: Reinforcing Large Language Models for Knowledge Base Question Answering
Xin Sun, Zhongqi Chen, Xing Zheng, Bowen Song, Qiang Liu, Shu Wu, Zilei Wang, Weiqiang Wang, Liang Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add relation retrieval and confidence gating (RRCG) (ablation) | KBQA-R1 · GrailQA | F1: 67.1 -> 86.1 | Improved |
| use multi-turn interaction instead of single-turn generation (ablation) | KBQA-R1 · GrailQA | F1: 49.8 -> 86.1 | Improved |
| use Referenced Rejection Sampling instead of standard rejection sampling (ablation) | KBQA-R1 warm-start · WebQSP | F1: 78.9 -> 83.4 | Improved |
| add SFT warm-start before RL (ablation) | KBQA-R1 · WebQSP | F1: 75.2 -> 83.4 | Improved |
| add GRPO RL after SFT (ablation) | KBQA-R1 · WebQSP | F1: 72.1 -> 83.4 | Improved |
| add format reward (ablation) | KBQA-R1 · GrailQA | F1: 84.2 -> 86.1 | Improved |
| use RRS rather than standard RS for trajectory collection (ablation) | SFT data synthesis · GrailQA | acceptance rate (%): 39.3 -> 67.0 | Improved |
| use RRS data instead of standard RS data for SFT init (ablation) | SFT initialization · GrailQA | SFT init F1: 73.8 -> 80.2 | Improved |
| give frontier models the KBQA-R1 structured harness instead of SPARQL ReAct | Claude 4.6 Sonnet · 500-query subset WebQSP | F1: 54.2 -> 76.7 | Improved |
| train learned RL policy instead of using harness with best frontier model | KBQA-R1 · 500-query subset average | F1: 71.6 -> 79.0 | Improved |
| GRPO training on relation grounding | KBQA-R1 · WebQSP | top-1 relation similarity: 0.9400 -> 0.9827 | Improved |
| KBQA-R1 vs DecAF on GrailQA I.I.D F1 | KBQA-R1 · GrailQA I.I.D | F1: 92.4 (DecAF) vs 91.5 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- KBQA-o1 — KBQA-o1: Agentic Knowledge Base Question Answering with Monte Carlo Tree Search (ICML 2025) · better on 5 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- KBQA-o1: Agentic Knowledge Base Question Answering with Monte Carlo Tree Search (ICML 2025 · similarity 0.92)
- Debate on Graph: A Flexible and Reliable Reasoning Framework for Large Language Models (AAAI 2025 · similarity 0.91)
- KG-Guard: Graph-Based Hallucination Detection for Knowledge Base Question Answering (NeurIPS 2026 · similarity 0.90)
- Promoting Knowledge Base Question Answering by Directing LLMs to Generate Task-relevant Logical Forms (AAAI 2025 · similarity 0.90)
- Cost-efficient Knowledge-based Question Answering with Large Language Models (NeurIPS 2024 · similarity 0.90)
- FlexKBQA: A Flexible LLM-Powered Framework for Few-Shot Knowledge Base Question Answering (AAAI 2024 · similarity 0.90)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
