Share your thoughts, 1 month free Claude Pro on usSee more

Few-shot Visual Question Answering on VL-ICL Bench Open-MI (test)

77.9Mean Accuracy

MAPD

Updated 1mo ago

Evaluation Results

Method	Links
MAPD 2025.06		77.9
Multi-TaskPD 2025.06		74.6
LoRA ([0-15] LLM layers + ATT) 2025.06		69.1
NoMeta-taskPD 2025.06		68
LoRA ([0-15] LLM layers) 2025.06		67.3
In-ContextPD 2025.06		64.5
Model-AvgPD 2025.06		63.1
LoRA (All LLM layers) 2025.06		55.1
MAPD 2025.06		53.3
In-ContextPD 2025.06		51.1
Multi-TaskPD 2025.06		48.6
NoMeta-taskPD 2025.06		43.8
Model-AvgPD 2025.06		26.6