Code Generation on Turbulence

Metric: CorrSc (higher is better)

LeaderboardDataset

Loading chart...

Results

Submit a result

Sort:

#	Model↕	CorrSc▼	Extra Data	Paper	Date↕	Code
1	GPT-4	0.848	No	Turbulence: Systematically and Automatically Tes...	2023-12-22	Code
2	GPT-3.5-Turbo	0.617	No	Turbulence: Systematically and Automatically Tes...	2023-12-22	Code
3	CodeLlama:13B-4bit-quantised	0.327	No	Turbulence: Systematically and Automatically Tes...	2023-12-22	Code
4	CodeLlama:7B-4bit-quantised	0.289	No	Turbulence: Systematically and Automatically Tes...	2023-12-22	Code
5	Command	0.063	No	Turbulence: Systematically and Automatically Tes...	2023-12-22	Code

#1GPT-4SOTA
0.848
CorrSc· 2023-12-22
Turbulence: Systematically and Automatically Testing Instruction-Tuned Large Language Models for Code Code
#2GPT-3.5-Turbo
0.617
CorrSc· 2023-12-22
Turbulence: Systematically and Automatically Testing Instruction-Tuned Large Language Models for Code Code
#3CodeLlama:13B-4bit-quantised
0.327
CorrSc· 2023-12-22
Turbulence: Systematically and Automatically Testing Instruction-Tuned Large Language Models for Code Code
#4CodeLlama:7B-4bit-quantised
0.289
CorrSc· 2023-12-22
Turbulence: Systematically and Automatically Testing Instruction-Tuned Large Language Models for Code Code
#5Command
0.063
CorrSc· 2023-12-22
Turbulence: Systematically and Automatically Testing Instruction-Tuned Large Language Models for Code Code