LLaVA-Bench
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
LLaVA-Bench COCO
20LLaVA-Bench Score
4
LLaVA-Bench In-the-Wild Detail
75.6Score
4
LLaVA-Bench In-the-Wild 1.0 (test)
58.8Conversation
4
LLaVA-Bench (eval)
0.971AUROC
3
Loading the SOTA2 catalog…