undefined | Better HN

0 pointskibwen2y ago0 comments

Unfortunately, no, because the existence of LLMs that can automatically determine code that is suspicious will be offset by the existence of LLMs that can generate malicious code that bypasses the detection abilities of the aforementioned LLMs.

0 comments

2 comments · 1 top-level

CrazyStat2y ago· 1 in thread

Generative Adversarial LLMs, let’s go!

protosam2y ago

Perhaps we could just call these ALLMs (Adversarial Large Language Models). You’re already dropping the N in GAN, I see no need for the G.

As an end result I think someone clever could make a LLaMA pun for the name of a LLaMA based ALLM.

j / k navigate · click thread line to collapse