Share your thoughts, 1 month free Claude Pro on usSee more

Code Generation on LiveCodeBench v5 (Pass@1)

61.5Pass@1

ESamp

Updated 1mo ago

Evaluation Results

Method	Links
ESamp 2026.04		61.5
OverRIDE 2026.04		57.6
Min-p 2026.04		52.5
Vanilla 2026.04		52.2
ESamp 2026.04		51.8
FIRE 2026.04		51.2
Min-p 2026.04		45.4
FIRE 2026.04		45.2
OverRIDE 2026.04		44.6
Vanilla 2026.04		43.4
PriFT-mass 2026.06		39.89
PriFT-prob 2026.06		39.32
Original 2026.06		30.45
ASFT 2026.06		28.64
IDFT 2026.06		26.14
TALR 2026.06		25
DFT 2026.06		24.32
EAFT 2026.06		23.86
ESamp 2026.04		22.6
SFT 2026.06		22.39
OverRIDE 2026.04		20.7
Min-p 2026.04		19.3
Vanilla 2026.04		19.2
FIRE 2026.04		18.3
OverRIDE 2026.04		13.6
OverRIDE 2026.04		13.6
Min-p 2026.04		13.5
Min-p 2026.04		13.5
ESamp 2026.04		13.1
ESamp 2026.04		13.1
Vanilla 2026.04		13
Vanilla 2026.04		13
FIRE 2026.04		12.9
FIRE 2026.04		12.9
ESamp 2026.04		12
FIRE 2026.04		11
OverRIDE 2026.04		10.3
Vanilla 2026.04		10
Min-p 2026.04		10