List your company

Visual Question Answering for object probing on POPE Aggregated random, popular, and adversarial

86.53Accuracy (POPE Aggregated)RUDDER-Beta
74.185277.390180.59583.7999Nov 13, 2025

Evaluation results

MethodLinks
RUDDER-BetaDecoding Strategy=Greedy, LVLM=LLaVA-1.52025.1186.5386.03
RUDDER-BetaDecoding Strategy=Beam Search, LVLM=LLaVA-1.52025.1186.5186.19
VISTADecoding Strategy=Greedy, LVLM=InstructBLIP2025.1186.2585.06
VISTADecoding Strategy=Greedy, LVLM=LLaVA-1.52025.1186.2185.42
VISTADecoding Strategy=Nucleus Sampling, LVLM=InstructBLIP2025.1186.1285.26
VISTADecoding Strategy=Beam Search, LVLM=LLaVA-1.52025.1186.185.35
RUDDER-AddDecoding Strategy=Greedy, LVLM=InstructBLIP2025.1186.0585.05
RUDDER-BetaDecoding Strategy=Greedy, LVLM=InstructBLIP2025.1186.0284.93
PAIDecoding Strategy=Greedy, LVLM=LLaVA-1.52025.1185.9885.31
RUDDER-AddDecoding Strategy=Beam Search, LVLM=LLaVA-1.52025.1185.9885.02
RUDDER-AddDecoding Strategy=Nucleus Sampling, LVLM=InstructBLIP2025.1185.9584.95
RUDDER-AddDecoding Strategy=Greedy, LVLM=LLaVA-1.52025.1185.9284.98
VCDDecoding Strategy=Greedy, LVLM=InstructBLIP2025.1185.7984.89
RUDDER-BetaDecoding Strategy=Nucleus Sampling, LVLM=InstructBLIP2025.1185.7984.74
VanillaDecoding Strategy=Greedy, LVLM=InstructBLIP2025.1185.7484.75
RUDDER-AddDecoding Strategy=Beam Search, LVLM=InstructBLIP2025.1185.7184.75
VISTADecoding Strategy=Beam Search, LVLM=InstructBLIP2025.1185.6484.61
VCDDecoding Strategy=Nucleus Sampling, LVLM=InstructBLIP2025.1185.6184.65
VCDDecoding Strategy=Beam Search, LVLM=LLaVA-1.52025.1185.685.06
PAIDecoding Strategy=Beam Search, LVLM=LLaVA-1.52025.1185.5885.01
RUDDER-BetaDecoding Strategy=Beam Search, LVLM=InstructBLIP2025.1185.5484.4
DoLaDecoding Strategy=Greedy, LVLM=LLaVA-1.52025.1185.5184.96
VanillaDecoding Strategy=Nucleus Sampling, LVLM=InstructBLIP2025.1185.584.52
VCDDecoding Strategy=Greedy, LVLM=LLaVA-1.52025.1185.4684.87
VanillaDecoding Strategy=Beam Search, LVLM=LLaVA-1.52025.1185.4684.98
VanillaDecoding Strategy=Greedy, LVLM=LLaVA-1.52025.1185.3484.91
VCDDecoding Strategy=Beam Search, LVLM=InstructBLIP2025.1184.9584.59
VanillaDecoding Strategy=Beam Search, LVLM=InstructBLIP2025.1184.7384.37
RUDDER-BetaDecoding Strategy=Nucleus Sampling, LVLM=LLaVA-1.52025.1184.0283.57
VISTADecoding Strategy=Nucleus Sampling, LVLM=LLaVA-1.52025.1183.5882.21
RUDDER-AddDecoding Strategy=Nucleus Sampling, LVLM=LLaVA-1.52025.1183.282.38
PAIDecoding Strategy=Nucleus Sampling, LVLM=LLaVA-1.52025.1183.1782.14
VanillaDecoding Strategy=Nucleus Sampling, LVLM=LLaVA-1.52025.118381.08
DoLaDecoding Strategy=Nucleus Sampling, LVLM=LLaVA-1.52025.1182.9481.12
VCDDecoding Strategy=Nucleus Sampling, LVLM=LLaVA-1.52025.1182.8281.9
RUDDER-BetaDecoding Strategy=Beam Search, LVLM=Idefics22025.1179.3377.96
RUDDER-AddDecoding Strategy=Beam Search, LVLM=Idefics22025.1178.9177.6
RUDDER-BetaDecoding Strategy=Greedy, LVLM=Idefics22025.1178.7476.52
VanillaDecoding Strategy=Beam Search, LVLM=Idefics22025.1178.6777.55
RUDDER-AddDecoding Strategy=Greedy, LVLM=Idefics22025.1178.4375.91
VanillaDecoding Strategy=Greedy, LVLM=Idefics22025.1178.474.86
VISTADecoding Strategy=Beam Search, LVLM=Idefics22025.1178.477.31
VISTADecoding Strategy=Greedy, LVLM=Idefics22025.1178.2874.66
RUDDER-BetaDecoding Strategy=Nucleus Sampling, LVLM=Idefics22025.1175.8969.69
RUDDER-AddDecoding Strategy=Nucleus Sampling, LVLM=Idefics22025.1174.9567.84
VanillaDecoding Strategy=Nucleus Sampling, LVLM=Idefics22025.1174.8467.78
VISTADecoding Strategy=Nucleus Sampling, LVLM=Idefics22025.1174.6667.7