List your company

Perception on TWI-oriented TreeBench (offline)

71.1AccuracyRTWI
62.7864.9467.169.26Feb 13, 2026

Evaluation results

MethodLinks
RTWIBase Model=Qwen3-VL Instruct, Setting=offline2026.0271.142
DeepconfBase Model=Qwen3-VL Instruct, Setting=offline2026.0269.838.9
Self-Cer.Base Model=Qwen3-VL Instruct, Setting=offline2026.0269.838.6
RTWIBase Model=Qwen3-VL Thinking, Setting=offline2026.0268.536
SCBase Model=Qwen3-VL Instruct, Setting=offline2026.0268.438.6
CISCBase Model=Qwen3-VL Instruct, Setting=offline2026.0268.438.6
SCBase Model=Qwen3-VL Thinking, Setting=offline2026.0266.432.5
CISCBase Model=Qwen3-VL Thinking, Setting=offline2026.0266.432.5
Self-Cer.Base Model=Qwen3-VL Thinking, Setting=offline2026.0265.832.5
DeepconfBase Model=Qwen3-VL Thinking, Setting=offline2026.0264.432.9
BaseBase Model=Qwen3-VL Instruct, Setting=offline2026.0264.4—
BaseBase Model=Qwen3-VL Thinking, Setting=offline2026.0263.1—