Research map / Large language models / Retrieval and knowledge
MedREK: Retrieval-Based Editing for Medical LLMs with Key-Aware Prompts
Shujun Xia, Haokun Lin, Yichen Wu, Yinan Zhou, Zixuan Li, Zhongwei Wan, Xingrun Xing, Yefeng Zheng, Xiang Li, Caifeng Shan, Zhenan Sun, Quanzheng Li
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 shared query-key MLP (ablation) | MedREK · MedCF++ 100 edits Meditron-7B | Avg: 69.91 -> 88.30 | Improved |
| add shared query-key MLP (ablation) | MedREK · MedCF++ 100 edits Meditron-7B | Efficacy: 56.74 -> 77.96 | Improved |
| add attention-based prompt encoder (ablation) | MedREK · MedCF++ 100 edits Meditron-7B | Avg: 85.24 -> 88.30 | Improved |
| add attention-based prompt encoder (ablation) | MedREK · MedCF++ 100 edits Meditron-7B | Generality: 75.63 -> 79.88 | Improved |
| use MedREK instead of RECIPE under 100-edit batch | retrieval-based editor · MedCF++ Meditron-7B | Avg: 71.83 -> 88.30 | Improved |
| use MedREK instead of RECIPE under 100-edit batch | retrieval-based editor · MedCF++ Meditron-7B | Efficacy: 68.30 -> 77.96 | Improved |
| use MedREK instead of RECIPE under 100-edit batch | retrieval-based editor · MedCF++ Meditron-7B | Generality: 70.74 -> 79.88 | Improved |
| increase batch size from 1 to 100 edits | RECIPE · MedCF++ Meditron-7B | Avg: 84.11 -> 71.83 | Hurt |
| increase batch size from 1 to 100 edits | MedREK · MedCF++ Meditron-7B | Avg: 89.33 -> 88.30 | No clear effect |
| use retrieval-based MedREK instead of parameter-modifying MedLaSA | editor · MedVersa single edit Meditron-7B | Avg: 79.97 -> 86.24 | Improved |
| use MedREK instead of MEMIT under 100 edits | editor · MedCF++ HuatuoGPT-o1-8B | Avg: 61.23 -> 86.83 | Improved |
| use MedREK instead of RECIPE on MedVersa single edit | editor · MedVersa Meditron-7B | Avg: 78.35 -> 86.24 | Improved |
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.
- MEND — MEND: Meta Demonstration Distillation for Efficient and Effective In-Context Learning (ICLR 2024) · better on 1 of 1 shared result
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
- MultiMedBench: A Scenario-Aware Benchmark for Evaluating Knowledge Editing in Medical VQA (AAAI 2026 · similarity 0.88)
- Medmarks: A Comprehensive Open-Source LLM Benchmark Suite for Medical Tasks (NeurIPS 2026 · similarity 0.88)
- MedJourney: Benchmark and Evaluation of Large Language Models over Patient Clinical Journey (NeurIPS 2024 · similarity 0.87)
- MedCalc-Bench: Evaluating Large Language Models for Medical Calculations (NeurIPS 2024 · similarity 0.85)
- MEMOIR: Lifelong Model Editing with Minimal Overwrite and Informed Retention for LLMs (NeurIPS 2025 · similarity 0.85)
- MediConfusion: Can you trust your AI radiologist? Probing the reliability of multimodal medical foundation models (ICLR 2025 · similarity 0.85)
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.
