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

Embodied Agent Task Execution on VAB-OmniGibson

27Success Rate (SR)

Benign SFT (GPT-4o)

Updated 4d ago

Evaluation Results

Method	Links
Benign SFT (GPT-4o) 2025.10		27	-	-	-
Original (GPT-4o) 2025.10		25	-	-	-
BEAT (InternVL3 8B) 2025.10		23	74.1	0	0.951
BEAT w/o CTL (GPT-4o) 2025.10		23	32.4	10	0.517
Benign SFT (InternVL3 8B) 2025.10		19	-	-	-
BEAT (Qwen2-VL 7B) 2025.10		18	77.9	0	0.923
Benign SFT (Qwen2-VL 7B) 2025.10		17	-	-	-
BEAT w/o CTL (InternVL3 8B) 2025.10		11	46.5	50	0.562
BEAT w/o CTL (Qwen2-VL 7B) 2025.10		10	47.6	7	0.713
Original (InternVL3 8B) 2025.10		1	-	-	-
Original (Qwen2-VL 7B) 2025.10		0	-	-	-