List your company

Visual Reasoning on VSP-Super

100Accuracy (Scale 16)EndoCoT
-4235077Mar 12, 2026

Evaluation results

MethodLinks
EndoCoTTraining Setting=Unified Training2026.0310080
DiffThinkerTraining Setting=Task-Specific Training2026.039980
EndoCoTTraining Setting=Task-Specific Training2026.039985
DiffThinkerTraining Setting=Unified Training2026.039784
Qwen3-VL-8BTraining Setting=Task-Specific Training, Optimization Method=SFT2026.03618
ChronoEditTraining Setting=Zero-Shot2026.0311
Qwen3-VL-8BTraining Setting=Zero-Shot2026.0310
ThinkGenTraining Setting=Zero-Shot2026.0300
Qwen-Image-Edit-2511Training Setting=Zero-Shot2026.0300
Qwen3-VL-8BTraining Setting=Task-Specific Training, Optimization Method=GRPO2026.0300