List your company

Medical Question Answering on Medical TF-QA (test)

85AccuracyPubMedBERT + MA-X
72.5275.767982.24Jan 19, 2026

Evaluation results

MethodLinks
PubMedBERT + MA-XArchitecture=Specialised Medical, Activation Steering=MEDASSESS-X2026.01858694
Gemma-3-27B + MA-XArchitecture=Decoder-Only, Activation Steering=MEDASSESS-X2026.018483115
Flan-T5-XL + MA-XArchitecture=Encoder-Decoder, Activation Steering=MEDASSESS-X2026.018382125
BioBERT + MA-XArchitecture=Specialised Medical, Activation Steering=MEDASSESS-X2026.018384103
BioGPT + MA-XArchitecture=Specialised Medical, Activation Steering=MEDASSESS-X2026.018382115
Llama-3-8B-Instruct + MA-XArchitecture=Decoder-Only, Activation Steering=MEDASSESS-X2026.018281136
T5-Large + MA-XArchitecture=Encoder-Decoder, Activation Steering=MEDASSESS-X2026.018180135
PubMedBERTArchitecture=Specialised Medical, Activation Steering=None2026.018180140
Mistral-7B-Instruct-v0.3 + MA-XArchitecture=Decoder-Only, Activation Steering=MEDASSESS-X2026.018079145
BioBERTArchitecture=Specialised Medical, Activation Steering=None2026.018079150
Gemma-3-27BArchitecture=Decoder-Only, Activation Steering=None2026.017976180
DeepSeek-7B + MA-XArchitecture=Decoder-Only, Activation Steering=MEDASSESS-X2026.017877165
Flan-T5-XLArchitecture=Encoder-Decoder, Activation Steering=None2026.017876190
BioGPTArchitecture=Specialised Medical, Activation Steering=None2026.017875170
Llama-3-8B-InstructArchitecture=Decoder-Only, Activation Steering=None2026.017774210
T5-LargeArchitecture=Encoder-Decoder, Activation Steering=None2026.017674200
Mistral-7B-Instruct-v0.3Architecture=Decoder-Only, Activation Steering=None2026.017572220
DeepSeek-7BArchitecture=Decoder-Only, Activation Steering=None2026.017370240