Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Agentic Task Success on MiniHack
Loading...
20
Success Rate
ALMA
6.168
9.759
13.35
16.941
Feb 8, 2026
Success Rate
Updated 1mo ago
Evaluation Results
Method
Method
Links
Success Rate
ALMA
Foundation Model=GPT-5...
2026.02
20
Trajectory Retrieval
Foundation Model=GPT-5...
2026.02
16.7
Reasoning Bank
Foundation Model=GPT-5...
2026.02
15.8
No Memory
Foundation Model=GPT-5...
2026.02
15
G-Memory
Foundation Model=GPT-5...
2026.02
14.2
ALMA
Foundation Model=GPT-5...
2026.02
11.7
Dynamic Cheatsheet
Foundation Model=GPT-5...
2026.02
11.7
Reasoning Bank
Foundation Model=GPT-5...
2026.02
9.8
Dynamic Cheatsheet
Foundation Model=GPT-5...
2026.02
9.2
Trajectory Retrieval
Foundation Model=GPT-5...
2026.02
7.5
G-Memory
Foundation Model=GPT-5...
2026.02
6.8
No Memory
Foundation Model=GPT-5...
2026.02
6.7
Feedback
Search any
task
Search any
task