List your company

Medical Question Answering on MedXpertQA

22.2AccuracyMA-RAG-ext
9.40812.72916.0519.371Feb 6, 2026Feb 14, 2026Feb 22, 2026Mar 3, 2026Mar 11, 2026Mar 19, 2026Mar 28, 2026

Evaluation results

MethodLinks
MA-RAG-extBackbone=Qwen3-8B, Ranking Agent=extrinsic verification2026.0222.2
MA-RAG-intBackbone=Qwen3-8B, Ranking Agent=intrinsic uncertainty2026.0221.2
IDCBackbone=Llama-3.1-8B, Method Type=Iterative Distractor Curation2026.0319.31
CoTBackbone=Qwen3-8B2026.0218
TC-RAGBackbone=Qwen3-8B2026.0218
FLAREBackbone=Qwen3-8B2026.0217.7
SR-RAGBackbone=Qwen3-8B2026.0217.3
FL-RAGBackbone=Qwen3-8B2026.0217
Citrus-VSize=8B2026.0316.9
FS-RAGBackbone=Qwen3-8B2026.0216.5
LingshuSize=7B2026.0316.5
MEDIC-ADSize=7B2026.0316.5
FilterBackbone=Llama-3.1-8B, Method Type=Filtering non-convertible questions2026.0316.45
Multi-RefineBackbone=Qwen3-8B2026.0216.3
Orig.Backbone=Llama-3.1-8B, Method Type=Standard RLVR on original MCQs2026.0316.29
Qwen3-8BBackbone=Qwen3-8B2026.0216.1
SCBackbone=Qwen3-8B2026.0215.7
RewriteBackbone=Llama-3.1-8B, Method Type=Model-based conversion2026.0315.67
HuatuoGPT-o1-8BBackbone=HuatuoGPT-o1-8B2026.0215.5
DirectBackbone=Llama-3.1-8B, Method Type=Direct conversion to short-answer2026.0315.47
BaseBackbone=Llama-3.1-8B, Method Type=Base instruct model2026.0314.29
UltraMedical-3.1-8BBackbone=UltraMedical-3.1-8B2026.0213.1
Llama-3.1-8BBackbone=Llama-3.1-8B2026.0212.4
IDCBackbone=Qwen2-7B, Method Type=Iterative Distractor Curation2026.0311.88
Orig.Backbone=Qwen2-7B, Method Type=Standard RLVR on original MCQs2026.0310.98
LingshuSize=7B, Evaluation Protocol=full finetuning2026.0310.9
FilterBackbone=Qwen2-7B, Method Type=Filtering non-convertible questions2026.0310.86
DirectBackbone=Qwen2-7B, Method Type=Direct conversion to short-answer2026.0310.29
BaseBackbone=Qwen2-7B, Method Type=Base instruct model2026.0310.24
RewriteBackbone=Qwen2-7B, Method Type=Model-based conversion2026.0310.24
LLaVA-MedSize=7B2026.039.9