List your company

Visual Reasoning on HR-Bench 4K FSP

96.5ACCRTWI
65.61273.63181.6589.669Feb 13, 2026

Evaluation results

MethodLinks
RTWIModel Variant=Qwen3-VL Instruct, Online setting=true2026.0296.566.2
DeepconfModel Variant=Qwen3-VL Instruct, Online setting=true2026.029660.3
ASCModel Variant=Qwen3-VL Instruct, Online setting=true2026.0295.859.8
CISCModel Variant=Qwen3-VL Instruct, Online setting=true2026.0295.557.7
SCModel Variant=Qwen3-VL Instruct, Online setting=true2026.0295.3—
ESCModel Variant=Qwen3-VL Instruct, Online setting=true2026.0295.360.7
Self-Cer.Model Variant=Qwen3-VL Instruct, Online setting=true2026.0295.362.5
BaseModel Variant=Qwen3-VL Instruct, Online setting=true2026.0293.8—
DeepEyesOnline setting=true2026.0291.3—
RTWIModel Variant=Qwen3-VL Thinking, Online setting=true2026.0291.361.9
Thyme2026.0291—
ThymeOnline setting=true2026.0291—
DeepconfModel Variant=Qwen3-VL Thinking, Online setting=true2026.0290.557.3
RTWIBackbone=DeepEyes-7B2026.029064.5
CISCBackbone=DeepEyes-7B2026.0289.856.4
SCBackbone=DeepEyes-7B2026.0289.5—
ASCBackbone=DeepEyes-7B2026.0289.558.9
ESCBackbone=DeepEyes-7B2026.0289.555.8
DeepconfBackbone=DeepEyes-7B2026.0289.553.9
Self-Uncer.Backbone=DeepEyes-7B2026.0289.561.9
SCModel Variant=Qwen3-VL Thinking, Online setting=true2026.0289.5—
ASCModel Variant=Qwen3-VL Thinking, Online setting=true2026.0289.551.9
ESCModel Variant=Qwen3-VL Thinking, Online setting=true2026.0289.540.5
Self-Cer.Model Variant=Qwen3-VL Thinking, Online setting=true2026.0289.547.6
CISCModel Variant=Qwen3-VL Thinking, Online setting=true2026.028832.3
Base*Backbone=DeepEyes-7B2026.0287.8—
BaseModel Variant=Qwen3-VL Thinking, Online setting=true2026.0283—
GPT-4o2026.0266.8—
GPT-4oOnline setting=true2026.0266.8—