When an LM's code fails a unit test, prompting the LM to explain why and suggesting changes enables it to generate improved versions (version 3) that work better, creating a feedback loop.

causalpending

Speaker

Andrew Ng

Evidence Quote

“if it fails a unit test then he why do you fail the unit test have that conversation and be able to figure out fail the unit test so you should try changing something and come up with V3”

Source

What's next for AI agentic workflows ft. Andrew Ng of AI Fund— Sequoia Capital
Created: 8/11/2026, 1:37:51 AM

Your notes

Loading notes...