HN
Hacker News
Top
New
Best
Ask
Show
Jobs
Comment by mschoening | Hacker News Reader
Parent
Full thread
mschoening
·
See Sleeper Agents (
https://arxiv.org/abs/2401.05566
).
View on HN
cosmojg
·
Who in their right mind is going to blindly take the code output by a large language model and toss it on a cruise missile? Sleeper agents are trivially circumvented by even a modicum of human oversight.
Reply on news.ycombinator.com