Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Software Engineering on SWE-bench Lite (300 instances)
Loading...
0
Misalignment Rate
Always Baseline
-2.2
12.65
27.5
42.35
May 27, 2026
Misalignment Rate
Solve Rate
Updated 6d ago
Evaluation Results
Method
Method
Links
Misalignment Rate
Solve Rate
Always Baseline
alpha (α)=0, Agent=Cla...
2026.05
0
12
CCO
alpha (α)=0.01, Agent=...
2026.05
1.3
13
CCO
alpha (α)=0.05, Agent=...
2026.05
5
14.3
CCO
alpha (α)=0.10, Agent=...
2026.05
9.7
15.3
CCO
alpha (α)=0.15, Agent=...
2026.05
14
16.7
CCO
alpha (α)=0.20, Agent=...
2026.05
17
17.7
CCO
alpha (α)=0.25, Agent=...
2026.05
21
19.7
Unconstrained
lambda (λ)=0, Agent=Cl...
2026.05
55
29
Feedback
Search any
task
Search any
task