Create: Update:
💻 AutoCodeRover: Autonomous Program Improvement
AutoCodeRover is a fully automated tool for fixing bugs on GitHub (fixing bugs in the issues section and generating new features for the project).
AutoCodeRover works in two stages:
🔎 Context search: LLM analyzes the code to collect context.
🔖 Patch generation: LLM rewrites code based on received context.
AutoCodeRover already solves ~16% of errors on the SWE-bench dataset and ~22% of errors in SWE-bench lite and continues to improve.
▪Github
▪Paper
#LLM
🆔 @Ai_Tv
AutoCodeRover is a fully automated tool for fixing bugs on GitHub (fixing bugs in the issues section and generating new features for the project).
AutoCodeRover works in two stages:
🔎 Context search: LLM analyzes the code to collect context.
🔖 Patch generation: LLM rewrites code based on received context.
AutoCodeRover already solves ~16% of errors on the SWE-bench dataset and ~22% of errors in SWE-bench lite and continues to improve.
▪Github
▪Paper
#LLM
🆔 @Ai_Tv
💻 AutoCodeRover: Autonomous Program Improvement
AutoCodeRover is a fully automated tool for fixing bugs on GitHub (fixing bugs in the issues section and generating new features for the project).
AutoCodeRover works in two stages:
🔎 Context search: LLM analyzes the code to collect context.
🔖 Patch generation: LLM rewrites code based on received context.
AutoCodeRover already solves ~16% of errors on the SWE-bench dataset and ~22% of errors in SWE-bench lite and continues to improve.
▪Github
▪Paper
#LLM
🆔 @Ai_Tv
AutoCodeRover is a fully automated tool for fixing bugs on GitHub (fixing bugs in the issues section and generating new features for the project).
AutoCodeRover works in two stages:
🔎 Context search: LLM analyzes the code to collect context.
🔖 Patch generation: LLM rewrites code based on received context.
AutoCodeRover already solves ~16% of errors on the SWE-bench dataset and ~22% of errors in SWE-bench lite and continues to improve.
▪Github
▪Paper
#LLM
🆔 @Ai_Tv
>>Click here to continue<<
هوش مصنوعی |یادگیری ماشین| علم داده