Hacker News Reader

How'd you do that?

I've described it pretty well have I not? Is there anything specific you would like to know?

> jailbreaking deepseek with a less aligned model (glm 5.2) which then allowed me to jailbreak deepseek with deepseek

Could you elaborate more on this specifically? What is the process to jailbreak one model with another?

I think the logic here is start a task with a model that won't refuse, then swap to a model that normally would. The model then sees a bunch of history that appears like its been partaking in the activity for a while.

My assumption is that since there's a history with that behavior, the model is less likely to question continuing it, since in most cases the "morality weights" would have triggered already at an earlier point. So in most cases, triggering it again would be a waste.

You are indeed correct, but also you can use a slightly less jailbroken model to build a stronger model, this is very effective on models that are of the same checkpoint as they're more likely to "feel" what makes them refuse.

Reply on news.ycombinator.com