Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Embodied AI Task Execution on EB-ALFRED online unsupervised setting
Loading...
61
Success Rate (Avg)
ELITE
25.64
34.82
44
53.18
Mar 25, 2026
Success Rate (Avg)
Success Rate (Base)
Success Rate (Long)
Updated 24d ago
Evaluation Results
Method
Method
Links
Success Rate (Avg)
Success Rate (Base)
Success Rate (Long)
ELITE
Backbone=Qwen2.5-VL-72...
2026.03
61
60
62
Qwen2.5-VL-72B-Ins
Model Type=Open-Source...
2026.03
52
55
49
ELITE
Backbone=InternVL3-78B
2026.03
44
40
48
InternVL2.5-78B
Model Type=Open-Source...
2026.03
40
38
42
ESCA
Backbone=Qwen2.5-VL-72...
2026.03
38
46
30
InternVL3-78B
Model Type=Open-Source...
2026.03
37
38
36
Qwen2-VL-72B-Ins
Model Type=Open-Source...
2026.03
35
40
30
gemma-3-27b-it
Model Type=Open-Source...
2026.03
34
42
26
Ovis2-34B
Model Type=Open-Source...
2026.03
29
34
24
Llama-3.2-90B-Vision-Ins
Model Type=Open-Source...
2026.03
27
38
16
Feedback
Search any
task
Search any
task