Instead of training an LLM to make comments, train the LLM to generate test case... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		lupire on Dec 22, 2024 \| parent \| context \| favorite \| on: How we made our AI code review bot stop leaving ni... Instead of training an LLM to make comments, train the LLM to generate test cases and then guess the line of code that breaks the test.

dakshgupta on Dec 22, 2024 [–]

You might enjoy @goodside on Twitter. He’s a prompt engineer at Scale AI and a lot of his observations and techniques are fascinating.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact