List your company

Hallucination Detection on AMBER sampled 5k

85.99A-ROCVIB-Probe
58.627665.731372.83579.9387Jan 9, 2026

Evaluation results

MethodLinks
VIB-ProbeBase VLM=LLaVA-v1.62026.0185.9985.91
VIB-ProbeBase VLM=Qwen2.5-VL2026.0185.5185.82
DHCPBase VLM=LLaVA-v1.62026.0184.7783.64
DHCPBase VLM=Qwen2.5-VL2026.0184.7783.98
VIB-ProbeBase VLM=Average2026.0184.4684.28
DHCPBase VLM=Average2026.0183.782.95
VIB-ProbeBase VLM=MiniGPT-42026.0183.482.94
DHCPBase VLM=MiniGPT-42026.0183.1882.27
VIB-ProbeBase VLM=LLaVA-v1.52026.0182.9582.43
DHCPBase VLM=LLaVA-v1.52026.0182.0781.89
RepProbingBase VLM=LLaVA-v1.52026.0177.5376.82
MetaTokenBase VLM=Qwen2.5-VL2026.0175.5975.1
RepProbingBase VLM=LLaVA-v1.62026.0175.3574.74
RepProbingBase VLM=Average2026.0174.9474.3
MetaTokenBase VLM=Average2026.0174.7174.63
RepProbingBase VLM=Qwen2.5-VL2026.0174.6174.52
MetaTokenBase VLM=LLaVA-v1.52026.0174.675.2
MetaTokenBase VLM=LLaVA-v1.62026.0174.4674.81
MetaTokenBase VLM=MiniGPT-42026.0174.1873.39
RepProbingBase VLM=MiniGPT-42026.0172.2571.11
AvgEntBase VLM=Qwen2.5-VL2026.0166.4266.8
AvgEntBase VLM=LLaVA-v1.62026.0165.4862.2
AvgProbBase VLM=LLaVA-v1.62026.0164.8163.79
AvgProbBase VLM=LLaVA-v1.52026.0164.7460.28
AvgProbBase VLM=Qwen2.5-VL2026.0164.5163.33
AvgEntBase VLM=Average2026.0163.862.49
AvgProbBase VLM=Average2026.0163.4460.83
AvgEntBase VLM=LLaVA-v1.52026.0162.0562.43
AvgEntBase VLM=MiniGPT-42026.0161.2558.53
AvgProbBase VLM=MiniGPT-42026.0159.6855.9