Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Code Generation on Wave-4 Wave-8 Qwen-2.5-7B (test)
Loading...
22.2
Achieved
Deployed ES
11.384
14.192
17
19.808
Jun 17, 2026
Achieved
Trace Oracle
Steps
Hindsight
Updated 1mo ago
Evaluation Results
Method
Method
Links
Achieved
Trace Oracle
Steps
Hindsight
Deployed ES
Threshold=T=peak+3
2026.06
22.2
-
51
49
MORBO-proxy
HPO=random Pareto HPO
2026.06
20.1
32
58
40.6
K-dim CARE
Gate=A3 + cpp_acc gate
2026.06
15.1
33.5
54
36.9
CARE v2
Gate Strategy=gate-adj...
2026.06
13.8
35
55
47.9
Naive
Budget=full budget
2026.06
11.8
34.2
50
37.8
Feedback
Search any
task
Search any
task