Perception on TWI-oriented TreeBench online setting
70.5AccuracyRTWI
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| RTWIModel Variant=Qwen3-VL Instruct, Setting=Online2026.02 | 70.5 | 40.8 | 50.7 | |
| SCModel Variant=Qwen3-VL Instruct, Setting=Online2026.02 | 68.4 | 38.6 | — | |
| ASCModel Variant=Qwen3-VL Instruct, Setting=Online2026.02 | 68.4 | 38.9 | 47.9 | |
| ESCModel Variant=Qwen3-VL Instruct, Setting=Online2026.02 | 68.4 | 38.7 | 40.9 | |
| DeepconfModel Variant=Qwen3-VL Instruct, Setting=Online2026.02 | 68.4 | 38.3 | 43.4 | |
| Self-Cer.Model Variant=Qwen3-VL Instruct, Setting=Online2026.02 | 68.4 | 38.9 | 48.5 | |
| RTWIModel Variant=Qwen3-VL Thinking, Setting=Online2026.02 | 67.8 | 35.1 | 60.1 | |
| CISCModel Variant=Qwen3-VL Instruct, Setting=Online2026.02 | 67.8 | 37.9 | 43.2 | |
| SCModel Variant=Qwen3-VL Thinking, Setting=Online2026.02 | 67.1 | 32.5 | — | |
| ASCModel Variant=Qwen3-VL Thinking, Setting=Online2026.02 | 67.1 | 32.1 | 40.1 | |
| ESCModel Variant=Qwen3-VL Thinking, Setting=Online2026.02 | 67.1 | 33.4 | 28 | |
| CISCModel Variant=Qwen3-VL Thinking, Setting=Online2026.02 | 67.1 | 32.9 | 49.7 | |
| DeepconfModel Variant=Qwen3-VL Thinking, Setting=Online2026.02 | 65.8 | 31.9 | 53.5 | |
| Self-Cer.Model Variant=Qwen3-VL Thinking, Setting=Online2026.02 | 64.4 | 33.1 | 38.9 | |
| BaseModel Variant=Qwen3-VL Instruct, Setting=Online2026.02 | 64.4 | — | — | |
| BaseModel Variant=Qwen3-VL Thinking, Setting=Online2026.02 | 63.1 | — | — |