Research map / Graph learning / Knowledge and temporal graphs
Conformal Path Reasoning: Trustworthy Knowledge Graph Question Answering via Path-Level Calibration
Shuhang Lin, Chuhao Zhou, Xiao Lin, Zihan Dong, Kuan Lu, Zhencan Peng, Jie Yin, Dimitris N. Metaxas
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 RCVNet learned path scoring to TreeG retrieval (ablation) | CPR · WebQSP (alpha=0.5) | ECR (%): 57.3 -> 61.0 | Improved |
| add PUCT-guided trajectory collection for RCVNet training (ablation) | CPR · WebQSP (alpha=0.5) | ECR (%): 61.0 -> 61.4 | Improved |
| add RCVNet (ablation) | CPR · PQL (alpha=0.5) | ECR (%): 51.4 -> 52.4 | Improved |
| add PUCT guidance (ablation) | CPR · PQL (alpha=0.5) | ECR (%): 52.4 -> 53.1 | Improved |
| add RCVNet (ablation) | CPR · WebQSP (alpha=0.8) | Coverage Efficiency: 0.200 -> 0.266 | Improved |
| add PUCT (ablation) | CPR · WebQSP (alpha=0.8) | Coverage Efficiency: 0.266 -> 0.314 | Improved |
| replace random-walk trajectories with 32 PUCT rollouts (ablation) | CPR · WebQSP (alpha=0.7) | Coverage Efficiency (%): 17.4 -> 18.3 | Improved |
| increase PUCT rollouts from 32 to 64 (ablation) | CPR · WebQSP (alpha=0.7) | Coverage Efficiency (%): 18.3 -> 18.9 | Improved |
| increase TreeG search budget (B,A) from (4,8) to (64,64) (ablation) | CPR · WebQSP (alpha=0.5) | ECR: ECR 0.582 -> 0.629; APSS 3.09 -> 9.02 | Mixed |
| add LLM-generated relation hints (Qwen3-8B vs none) (ablation) | CPR · WebQSP (alpha=0.8) | ECR (%): 42.0 -> 50.8 | Improved |
| add LLM-generated relation hints (Qwen3-8B vs none) (ablation) | UaG · WebQSP (alpha=0.8) | ECR (%): 19.8 -> 25.6 | Improved |
| replace hop-level calibration (UaG) with path-level query calibration (CPR) | CPR vs UaG · WebQSP (alpha=0.3) | APSS: 98.81 -> 17.77 | Improved |
Closest papers by meaning
- Diagnosing and Addressing Pitfalls in KG-RAG Datasets: Toward More Reliable Benchmarking (NeurIPS 2025 · similarity 0.89)
- KG-Guard: Graph-Based Hallucination Detection for Knowledge Base Question Answering (NeurIPS 2026 · similarity 0.89)
- EPERM: An Evidence Path Enhanced Reasoning Model for Knowledge Graph Question and Answering (AAAI 2025 · similarity 0.88)
- Debate on Graph: A Flexible and Reliable Reasoning Framework for Large Language Models (AAAI 2025 · similarity 0.88)
- TrustUQA: A Trustful Framework for Unified Structured Data Question Answering (AAAI 2025 · similarity 0.87)
- Graph Reasoning Transformers for Knowledge-Aware Question Answering (AAAI 2024 · similarity 0.87)
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.
