List your company

Multiple Choice VQA Benchmarks Average

84.2AUCCOMET
16.49634.07351.6569.227Jun 30, 2026

Evaluation results

MethodLinks
COMETBackbone=Qwen3VL-2B-Instruct, Generation-free (GF)=true2026.0684.292.943.8
VL UncertaintyBackbone=Qwen3VL-2B-Instruct, Generation-free (GF)=false2026.0680.290.948.2
COMETBackbone=Qwen3VL-8B-Instruct, Generation-free (GF)=true2026.0678.382.565.4
COMETBackbone=Qwen3VL-4B-Instruct, Generation-free (GF)=true2026.0678.284.365.5
Consist. (Ent)Backbone=Qwen3VL-2B-Instruct, Generation-free (GF)=false2026.0677.589.956.2
Consist. (Max)Backbone=Qwen3VL-2B-Instruct, Generation-free (GF)=false2026.0676.689.358.6
P(Correct)Backbone=Qwen3VL-8B-Instruct, Generation-free (GF)=false2026.0673.377.172.4
P(Correct)Backbone=Qwen3VL-4B-Instruct, Generation-free (GF)=false2026.0672.977.676.3
VL UncertaintyBackbone=Qwen3VL-8B-Instruct, Generation-free (GF)=false2026.0672.278.968.7
VL UncertaintyBackbone=Qwen3VL-4B-Instruct, Generation-free (GF)=false2026.0672.181.266
Candidate (Ent)Backbone=Qwen3VL-2B-Instruct, Generation-free (GF)=true2026.0671.980.179.9
Consist. (Ent)Backbone=Qwen3VL-8B-Instruct, Generation-free (GF)=false2026.0671.479.165.8
Consist. (Ent)Backbone=Qwen3VL-4B-Instruct, Generation-free (GF)=false2026.0670.981.363.8
Consist. (Max)Backbone=Qwen3VL-8B-Instruct, Generation-free (GF)=false2026.0670.978.168.8
Candidate (Max)Backbone=Qwen3VL-2B-Instruct, Generation-free (GF)=true2026.0670.879.381.2
Consist. (Max)Backbone=Qwen3VL-4B-Instruct, Generation-free (GF)=false2026.0670.680.767.8
Candidate (Ent)Backbone=Qwen3VL-4B-Instruct, Generation-free (GF)=true2026.066867.686.1
Consist. (Ent)Backbone=InternVL3.5-1B, Generation-free (GF)=false2026.066880.476.7
Candidate (Ent)Backbone=Qwen3VL-8B-Instruct, Generation-free (GF)=true2026.0667.564.684.6
Consist. (Max)Backbone=InternVL3.5-1B, Generation-free (GF)=false2026.0667.479.878.1
Candidate (Max)Backbone=Qwen3VL-4B-Instruct, Generation-free (GF)=true2026.0667.366.488
Candidate (Max)Backbone=Qwen3VL-8B-Instruct, Generation-free (GF)=true2026.0666.963.586.7
COMETBackbone=InternVL3.5-1B, Generation-free (GF)=true2026.066677.285.1
Verbalized ConfBackbone=Qwen3VL-2B-Instruct, Generation-free (GF)=false2026.0661.777.393.3
P(Correct)Backbone=InternVL3.5-1B, Generation-free (GF)=false2026.0658.971.691.1
VL UncertaintyBackbone=InternVL3.5-1B, Generation-free (GF)=false2026.0658.771.590.2
Candidate (Ent)Backbone=InternVL3.5-1B, Generation-free (GF)=true2026.0658.365.792.1
Candidate (Max)Backbone=InternVL3.5-1B, Generation-free (GF)=true2026.0657.565.192.8
Verbalized ConfBackbone=Qwen3VL-8B-Instruct, Generation-free (GF)=false2026.06566489.9
Verbalized ConfBackbone=Qwen3VL-4B-Instruct, Generation-free (GF)=false2026.065467.691.1
Verbalized ConfBackbone=InternVL3.5-1B, Generation-free (GF)=false2026.0653.169.589.2
P(Correct)Backbone=Qwen3VL-2B-Instruct, Generation-free (GF)=false2026.0651.168.897
NLLBackbone=InternVL3.5-1B, Generation-free (GF)=false2026.0635.755.899.4
NLLBackbone=Qwen3VL-8B-Instruct, Generation-free (GF)=false2026.0630.145.298
NLLBackbone=Qwen3VL-4B-Instruct, Generation-free (GF)=false2026.0624.744.699.6
NLLBackbone=Qwen3VL-2B-Instruct, Generation-free (GF)=false2026.0619.155.699.6