List your company

Multimodal VQA on LLaVA 1.5 Subset (GQA, MMB, POPE, MME) (test)

100AccuracyVanilla (LLaVA)
75.87282.13688.494.664Jun 14, 2026

Evaluation results

MethodLinks
Vanilla (LLaVA)Tokens=576, Backbone=LLaVA-v1.5-7B, Hardware=NVIDIA RTX A60002026.061008.89149.51
StepPruneTokens=64, K=64, Backbone=LLaVA-v1.5-7B, Hardware=NVIDIA RTX A60002026.0694.62.2879.51.88
PDropTokens=64, K=64, Backbone=LLaVA-v1.5-7B, Hardware=NVIDIA RTX A60002026.0682.74.89105.91.41
FastVTokens=64, K=64, Backbone=LLaVA-v1.5-7B, Hardware=NVIDIA RTX A60002026.0676.82.39102.61.46