Our new X account is live! Follow @wizwand_team for updates

Multi-modal Reasoning on EMMA

32.7Accuracy

GPT-4o

Updated 3d ago

Evaluation Results

Method	Links
GPT-4o 2026.01		32.7
ReLaX-VL-7B 2025.12		30.6
VL-Rethinker-7B 2025.12		29.7
SRPO-7B 2025.12		29.6
DAPO (Ours) 2026.01		29.3
Two-stage RL (Ours) 2026.01		28.6
DPS (Ours) 2026.01		28.4
ReLaX-VL-3B 2025.12		26.9
OpenVLThinker-7B 2025.12		26.6
mPLUG-Owl3 2026.01		24.8
MM-Eureka-7B 2025.12		23.5
TW-GRPO 2026.01		22.5
Vision-R1-7B 2025.12		22.4
Qwen2.5-VL-7B 2025.12		21.5
MM-Eureka-8B 2025.12		21.5
InternVL2.5 2026.01		21
Intern2.5-VL-8B 2025.12		20.6
Qwen2.5-VL 2026.01		20.4
Mantis-Idefics2 2026.01		20.3
Qwen2-VL-7B 2025.12		20.2
Intern2-VL-8B 2025.12		19.8
Qwen2.5-VL-3B 2025.12		19.2
LLaVA-NeXT-Interleave 2026.01		19
Llava-OV-7B 2025.12		18.3
VideoRFT 2026.01		17.8
R1-VL-7B 2025.12		8.3