Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Task Success on ALFWorld (50 eval tasks)
Loading...
49
Success Rate
TraceRetain-CEM-K50
38.6
41.3
44
46.7
Jun 28, 2026
Success Rate
Average Steps
Precision@5
Updated 26d ago
Evaluation Results
Method
Method
Links
Success Rate
Average Steps
Precision@5
TraceRetain-CEM-K50
Mem=50
2026.06
49
9.52
29.2
TraceRetain-Linear-K50
Mem=50
2026.06
48
12.1
30
FIFO-K50
Mem=50
2026.06
48
11.08
26.4
Unbounded
Mem=100
2026.06
47
10.82
45.2
No memory
Mem=0
2026.06
39
20.98
0
Feedback
Search any
task
Search any
task