List your company

Multiple Choice Question Answering on MedQA

50.98AccuracyDS2-INSTRUCT
12.55222.528532.50542.4815Mar 30, 2025May 27, 2025Jul 24, 2025Sep 20, 2025Nov 17, 2025Jan 14, 2026Mar 13, 2026

Evaluation results

MethodLinks
DS2-INSTRUCTModel Family=Qwen2.52026.0350.98—
Self-InstructModel Family=Qwen2.52026.0348.92—
Zero-ShotModel Family=Qwen2.52026.0348.35—
InstructMixModel Family=Qwen2.52026.0347.83—
ExploreInstructModel Family=Qwen2.52026.0347.15—
QAPBackbone=Flan-T5, Parameter Count=11B2025.0344.010.34
DS2-INSTRUCTModel Family=Mistral2026.0342.99—
QAPBackbone=Llama2-chat, Parameter Count=13B2025.0342.560.34
PEBackbone=Llama2-chat, Parameter Count=13B2025.0340.830.1
QAPBackbone=Llama2-chat, Parameter Count=7B2025.0339.940.33
GNPBackbone=Flan-T5, Parameter Count=11B2025.0339.760.31
Self-InstructModel Family=Mistral2026.0339.74—
PEBackbone=Flan-T5, Parameter Count=11B2025.0339.540.15
LLMBackbone=Flan-T5, Parameter Count=11B2025.0339.180.07
DS2-INSTRUCTModel Family=Llama32026.0339.15—
LLMBackbone=Llama2-chat, Parameter Count=13B2025.0339.060.14
SPBackbone=Flan-T5, Parameter Count=11B2025.0338.980.33
QAPBackbone=Flan-T5, Parameter Count=3B2025.0338.330.3
ExploreInstructModel Family=Mistral2026.0337.92—
LLMBackbone=Llama2-chat, Parameter Count=7B2025.0336.780.17
PEBackbone=Llama2-chat, Parameter Count=7B2025.0336.330.17
KGEPBackbone=Llama2-chat, Parameter Count=7B2025.0335.180.39
PEBackbone=Flan-T5, Parameter Count=3B2025.0334.360.11
LLMBackbone=Flan-T5, Parameter Count=3B2025.0334.220.14
SPBackbone=Flan-T5, Parameter Count=3B2025.0333.980.49
KGEPBackbone=Flan-T5, Parameter Count=11B2025.0333.930.46
GNPBackbone=Flan-T5, Parameter Count=3B2025.0333.870.26
KGEPBackbone=Llama2-chat, Parameter Count=13B2025.0333.370.29
InstructMixModel Family=Llama32026.0333.29—
GNPBackbone=Llama2-chat, Parameter Count=7B2025.0332.010.29
Self-InstructModel Family=Llama32026.0331.58—
ExploreInstructModel Family=Llama32026.0330.46—
KGEPBackbone=Flan-T5, Parameter Count=3B2025.0330.210.41
InstructMixModel Family=Mistral2026.0330.15—
SPBackbone=Llama2-chat, Parameter Count=7B2025.0327.820.28
Zero-ShotModel Family=Llama32026.0327.38—
GNPBackbone=Llama2-chat, Parameter Count=13B2025.0325.010.28
SPBackbone=Llama2-chat, Parameter Count=13B2025.0323.940.22
Zero-ShotModel Family=Mistral2026.0314.03—