List your company

Multimodal Reasoning on Multimodal Reasoning Aggregate 6 benchmarks

80.8Average ScoreM2Note
72.32474.524576.72578.9255Jul 1, 2026

Evaluation results

MethodLinks
M2NoteBackbone=Qwen3-VL-Plus, Evolving Mode=Self-Evolving, Mem=28, Len=4622026.0780.8
M2NoteBackbone=GPT-5.4, Evolving Mode=Self-Evolving, Mem=28, Len=4622026.0779.6
VanillaBackbone=Qwen3-VL-Plus, Evolving Mode=Self-Evolving2026.0779.5
VanillaBackbone=GPT-5.4, Evolving Mode=Self-Evolving2026.0778.8
CoT+M2NoteBackbone=Qwen3-VL-8B-Instruct, Evolving Mode=Self-Evolving, Mem=32, Len=3152026.0778.4
CoTBackbone=Qwen3-VL-8B-Instruct, Evolving Mode=Self-Evolving2026.0777.6
M2NoteBackbone (Tuning)=Qwen3-VL-8B-Instruct, Backbone (Tuner)=Qwen3-VL-Plus, Evolving Mode=Cross-Model Evolving, Mem=43, Len=4452026.0776.6
M2NoteBackbone=Qwen3-VL-8B-Instruct, Evolving Mode=Self-Evolving, Mem=50, Len=2612026.0776.2
VanillaBackbone=Qwen3-VL-8B-Instruct, Evolving Mode=Self-Evolving2026.0774.3
VanillaBackbone (Tuning)=Qwen3-VL-8B-Instruct, Backbone (Tuner)=Qwen3-VL-Plus, Evolving Mode=Cross-Model Evolving2026.0774.3
Vision-ZeroEvolving Mode=Self-Evolving2026.0772.65