List your company

Hallucination Detection on POPE official (val)

99.13A-PRLyapunov Probes
61.960471.610281.2690.9098Jan 9, 2026Jan 18, 2026Jan 27, 2026Feb 6, 2026Feb 15, 2026Feb 24, 2026Mar 6, 2026

Evaluation results

MethodLinks
Lyapunov ProbesModel=LLaVA-1.52026.0399.13—
Lyapunov ProbesModel=Qwen-2.5-VL2026.0399—
ProbeModel=Qwen-2.5-VL2026.0398.41—
ProbeModel=LLaVA-1.52026.0398.08—
VIB-ProbeBase VLM=LLaVA-v1.52026.0196.9696.52
DHCPBase VLM=Qwen2.5-VL2026.0196.5296.8
VIB-ProbeBase VLM=Qwen2.5-VL2026.0196.496.98
RepProbingBase VLM=Qwen2.5-VL2026.0195.8996.82
VIB-ProbeBase VLM=Average2026.0195.5695.92
VIB-ProbeBase VLM=LLaVA-v1.62026.0195.5195.99
MetaTokenBase VLM=Qwen2.5-VL2026.0194.8494.1
DHCPBase VLM=LLaVA-v1.52026.0194.5394.87
RepProbingBase VLM=LLaVA-v1.52026.0194.594.68
MetaTokenBase VLM=LLaVA-v1.62026.0194.3394.21
DHCPBase VLM=Average2026.0194.2595.1
DHCPBase VLM=LLaVA-v1.62026.0194.294.92
RepProbingBase VLM=Average2026.0194.1493.92
RepProbingBase VLM=LLaVA-v1.62026.0193.8793.01
VIB-ProbeBase VLM=MiniGPT-42026.0193.3794.19
MetaTokenBase VLM=Average2026.0192.8792.77
RepProbingBase VLM=MiniGPT-42026.0192.391.18
MetaTokenBase VLM=LLaVA-v1.52026.0192.2293.07
DHCPBase VLM=MiniGPT-42026.0191.7693.8
MetaTokenBase VLM=MiniGPT-42026.0190.0789.69
AvgEntBase VLM=LLaVA-v1.62026.0170.6679.51
AvgEntBase VLM=Qwen2.5-VL2026.0170.278.99
AvgEntBase VLM=Average2026.0169.2678.05
AvgEntBase VLM=MiniGPT-42026.0168.6476.27
AvgEntBase VLM=LLaVA-v1.52026.0167.5277.43
AvgProbBase VLM=LLaVA-v1.62026.0166.4463.06
AvgProbBase VLM=Average2026.0164.5764.29
AvgProbBase VLM=Qwen2.5-VL2026.0164.5568.28
AvgProbBase VLM=LLaVA-v1.52026.0163.964.25
AvgProbBase VLM=MiniGPT-42026.0163.3961.56