List your company
ResearchFormerly WizWand

Action-relation hallucination evaluation on R-Bench Instance

75.86AccuracyRVE
58.554463.047267.5472.0328May 12, 2026
Updated 5mo ago

Evaluation results

MethodLinks
RVELVLM Model=ShareGPT4V-7B2026.0575.8686
RVELVLM Model=LLaVA-1.5-13B2026.0575.7385.6
RVELVLM Model=LLaVA-1.5-7B2026.0574.4684.97
RVELVLM Model=LLaVA-NeXT-7B2026.0574.4684.56
VAFLVLM Model=LLaVA-NeXT-7B2026.0573.2283.35
VCDLVLM Model=LLaVA-NeXT-7B2026.0572.8783.03
RVELVLM Model=InstructBLIP-7B2026.0570.6481.2
VAFLVLM Model=LLaVA-1.5-13B2026.0570.4580.67
VCDLVLM Model=LLaVA-1.5-13B2026.0570.0580.43
VCDLVLM Model=ShareGPT4V-7B2026.0569.7779.77
VanillaLVLM Model=LLaVA-1.5-13B2026.0569.1279.37
ICDLVLM Model=LLaVA-1.5-13B2026.0569.0179.3
VAFLVLM Model=ShareGPT4V-7B2026.0568.9579.1
VCDLVLM Model=LLaVA-1.5-7B2026.0568.6579.76
ICDLVLM Model=InstructBLIP-7B2026.0568.5379.33
ICDLVLM Model=LLaVA-NeXT-7B2026.0568.578.8
VAFLVLM Model=InstructBLIP-7B2026.0568.0579
VAFLVLM Model=LLaVA-1.5-7B2026.0567.6977.61
VCDLVLM Model=InstructBLIP-7B2026.0567.378.15
VanillaLVLM Model=InstructBLIP-7B2026.0566.777.57
VanillaLVLM Model=ShareGPT4V-7B2026.0566.3776.54
ICDLVLM Model=ShareGPT4V-7B2026.0566.2576.4
VanillaLVLM Model=LLaVA-NeXT-7B2026.0564.5983.15
VanillaLVLM Model=LLaVA-1.5-7B2026.0562.9473.45
ICDLVLM Model=LLaVA-1.5-7B2026.0559.2269.83