STEM Puzzle on MathVista
83.9AccuracyCoT+M2Note
Evaluation Results
| Method | Links | |
|---|---|---|
| CoT+M2NoteBackbone=Qwen3-VL-8B-Instruct, Evolving Mode=Self-Evolving, Mem=32, Len=3152026.07 | 83.9 | |
| CoTBackbone=Qwen3-VL-8B-Instruct, Evolving Mode=Self-Evolving2026.07 | 82.9 | |
| M2NoteBackbone=Qwen3-VL-Plus, Evolving Mode=Self-Evolving, Mem=28, Len=4622026.07 | 80.1 | |
| M2NoteBackbone (Tuning)=Qwen3-VL-8B-Instruct, Backbone (Tuner)=Qwen3-VL-Plus, Evolving Mode=Cross-Model Evolving, Mem=43, Len=4452026.07 | 78.3 | |
| VanillaBackbone=Qwen3-VL-Plus, Evolving Mode=Self-Evolving2026.07 | 78 | |
| M2NoteBackbone=GPT-5.4, Evolving Mode=Self-Evolving, Mem=28, Len=4622026.07 | 78 | |
| M2NoteBackbone=Qwen3-VL-8B-Instruct, Evolving Mode=Self-Evolving, Mem=50, Len=2612026.07 | 77.3 | |
| VanillaBackbone=GPT-5.4, Evolving Mode=Self-Evolving2026.07 | 77.2 | |
| DPEBackbone=Qwen3-VL-8B-Instruct, Evolving Mode=Self-Evolving2026.07 | 76.2 | |
| VanillaBackbone=Qwen3-VL-8B-Instruct, Evolving Mode=Self-Evolving2026.07 | 73.2 | |
| VanillaBackbone (Tuning)=Qwen3-VL-8B-Instruct, Backbone (Tuner)=Qwen3-VL-Plus, Evolving Mode=Cross-Model Evolving2026.07 | 73.2 | |
| Vision-ZeroEvolving Mode=Self-Evolving2026.07 | 72.6 | |
| EvoLMMEvolving Mode=Self-Evolving2026.07 | 70.5 | |
| iReasonerEvolving Mode=Self-Evolving2026.07 | 69.7 | |
| VisPlayEvolving Mode=Self-Evolving2026.07 | 68.2 |