Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Embodied Planning on ALFRED
Loading...
45.81
Success Rate (SR)
LoTA (Full Recompute)
15.2756
23.2028
31.13
39.0572
Feb 27, 2026
Success Rate (SR)
Time To First Fixation (s)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Success Rate (SR)
Time To First Fixation (s)
LoTA (Full Recompute)
Model=Qwen-2.5-32B (INT4)
2026.02
45.81
1.213
LoTA + KEEP
Model=Qwen-2.5-32B (INT4)
2026.02
45.5
0.635
LoTA (Full Recompute)
Model=Qwen-2.5-14B
2026.02
44.63
0.41
LoTA + KEEP
Model=Qwen-2.5-14B
2026.02
44.3
0.236
KARMA
Model=GPT-4o
2026.02
43
-
LoTA + CacheBlend
Model=Qwen-2.5-32B (INT4)
2026.02
41.37
1.209
FLARE
Model=GPT-4
2026.02
40.05
-
LoTA + CacheBlend
Model=Qwen-2.5-14B
2026.02
39.36
0.363
LoTA + Full Reuse
Model=Qwen-2.5-32B (INT4)
2026.02
35.72
0.602
LoTA + Full Reuse
Model=Qwen-2.5-14B
2026.02
34.41
0.21
LLM-Planner
Model=GPT-3
2026.02
16.45
-
Feedback
Search any
task
Search any
task