Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Code Generation on LeetCode without write access
Loading...
15.2
Pass@1
No intervention
0.64
4.42
8.2
11.98
Apr 1, 2026
Pass@1
Updated 16d ago
Evaluation Results
Method
Method
Links
Pass@1
No intervention
Model=Phi-4-mini, Writ...
2026.04
15.2
Adv. Mod. (multiplicative)
Model=Phi-4-mini, Alph...
2026.04
15
Adv. Mod. (additive)
Model=Phi-4-mini, Alph...
2026.04
12.7
Gen.-time suppression
Model=Phi-4-mini, Writ...
2026.04
12.6
Adv. Mod. (multiplicative)
Model=Phi-4-mini, alph...
2026.04
12
No intervention
Model=Llama-3.2-3B, Wr...
2026.04
6.8
Adv. Mod. (multiplicative)
Model=Llama-3.2-3B, Al...
2026.04
6.8
Gen.-time suppression
Model=Llama-3.2-3B, Wr...
2026.04
6.6
Adv. Mod. (additive)
Model=Llama-3.2-3B, Al...
2026.04
6.4
Gen.-time suppression
Model=Phi-4-mini
2026.04
6.1
Adv. Mod. (multiplicative)
Model=Llama-3.2-3B, al...
2026.04
5.1
Adv. Mod. (additive)
Model=Phi-4-mini, alph...
2026.04
4.2
Gen.-time suppression
Model=Llama-3.2-3B
2026.04
3
Adv. Mod. (additive)
Model=Llama-3.2-3B, al...
2026.04
3
No intervention
Model=Llama-3.2-3B
2026.04
1.9
No intervention
Model=Phi-4-mini
2026.04
1.2
Feedback
Search any
task
Search any
task