How to Backdoor Large Language Models

[jonasrosland]

Now this is a bit scary. Shrivu Shankar writes about how he trained an LLM to insert malicious backdoors into code it writes.

How to Backdoor Large Language Models