Research map / Large language models / Reasoning and chain of thought
CRAF: A Clinical Reasoning-Adaptive Framework via Reinforcement Learning for Similar Case Retrieval
Jie Lin, Lei Jiang, Zongyi Chen, Liansheng Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes CRAF, a reinforcement-learning framework that adapts clinical reasoning for similar case retrieval, with explicit chain-of-thought to surface implicit reasoning cues in clinical data.
Problem
Similar clinical case retrieval is hard because implicit reasoning cues in clinical data prevent current retrieval models from capturing deep semantic relations.
Approach
Applies reinforcement learning to reasoning-based retrieval so that explicit chain-of-thought generation is adapted to clinical case retrieval; details beyond the abstract are not available.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add CRAF-8B query rewriting | Qwen3-Embedding retriever · PMC-Patients | AVG: 36.33 -> 39.87 | Improved |
| add CRAF-8B query rewriting | BM25 · PMC-Patients | AVG: 27.42 -> 34.35 | Improved |
| train Qwen3-8B with CRAF RL rather than use it untrained for rewriting | Qwen3-8B rewriter + Qwen3-Embedding · PMC-Patients | AVG: 36.64 -> 39.87 | Improved |
| rewrite queries with HuatuoGPT-o1 for dense retrieval | Qwen3-Embedding · PMC-Patients | AVG: 36.33 -> 35.11 | Hurt |
| rewrite queries with CRAF-8B | Qwen3-Embedding · ReCDS-PPR | MRR: 27.05 -> 28.33 | Mixed |
| train with RL reward instead of SFT (ablation) | CRAF · ReCDS-PAR | retrieval metrics | Improved |
| add negative sampling and reward components (ablation) | CRAF · ReCDS-PAR | retrieval metrics | Improved |
| increase training data from 0 to 100 percent (ablation) | CRAF · ReCDS-PAR | retrieval metrics | Mixed |
Closest papers by meaning
- CLR-voyance : Reinforcing Open-Ended Reasoning for Inpatient Clinical Decision Support with Outcome-Aware Rubrics (NeurIPS 2026 · similarity 0.86)
- Large Language Models Are Clinical Reasoners: Reasoning-Aware Diagnosis Framework with Prompt-Generated Rationales (AAAI 2024 · similarity 0.86)
- Reasoning-Enhanced Healthcare Predictions with Knowledge Graph Community Retrieval (ICLR 2025 · similarity 0.85)
- Simulating Viva Voce Examinations to Evaluate Clinical Reasoning in Large Language Models (NeurIPS 2025 · similarity 0.85)
- MedS³: Towards Medical Slow Thinking with Self-Evolved Soft Dual-sided Process Supervision (AAAI 2026 · similarity 0.85)
- MedReasoner: Reinforcement Learning Drives Reasoning Grounding from Clinical Thought to Pixel-Level Precision (AAAI 2026 · similarity 0.84)
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.
