Embedded Files
Main results on old LLMs including CodeGen, Incoder and SantacoderÂ
Main results on old LLMs including CodeGen, Incoder and SantacoderÂ
Performance
Detection performance for syntactically correct/incorrect code
Detection performance for syntactically correct/incorrect code
Supplementary results in RQ4: Limitations in Uncertainty Estimation for Code Generation
Supplementary results in RQ4: Limitations in Uncertainty Estimation for Code Generation
Partial correct code v.s. Correct code
Partial correct code v.s. Correct code
Detection Performance (AUC Score)
Detection Performance (AUC Score)
Syntactically Incorrect code v.s. Correct code
Syntactically Incorrect code v.s. Correct code
Detection Performance (AUC Score)
Detection Performance (AUC Score)
Note: GPT3.5, GPT4o mini and GPT4o do not have Syntactically Incorrect codeÂ
Page updated
Google Sites
Report abuse