undefined | Better HN

0 pointsipython3d ago0 comments

As far as I know the only data of the two you identified are cached inside of the inference layer - the KV cache. Then again, I am not an expert in designing and operating inference, so I could be incorrect on that.

Either way, both of those are controlled by deterministic code and not the LLM itself. So controlling for that risk is much simpler to model IMO since the mitigation can be applied universally and deterministically rather than hoping and praying some non-deterministic system will respect your wishes.

0 comments

3 comments · 1 top-level

wolttam3d ago· 2 in thread

In other words: controlling for that kind of potential data-mixing is the same as in any other application where customer data is co-located within the same running process/memory/storage space.

jcgrillo3d ago

Yes, however the companies that are responsible for doing it have already shown their asses a little bit with all the jailbreaking stuff, and we know they produce really awful code from all the recent harness issues... To my mind that indicates this critical invariant deserves a little scrutiny. But with all the vibe slop being slung these days who knows what's safe anymore.

All that is to say I sure would appreciate a coherent, clear technical explanation of how they ensure user data are separate while serving concurrent queries.

wolttam3d ago

They’re valid things to be concerned about IMO.

I think you’re looking for an answer you’re not going to get unfortunately. I think there actually is a higher than average risk of data leakage with the insane optimizations that go into model serving - GLM5.1 had an issue of going into jibberish when their infra was under high load, and it turned out to be a cross-request KV cache contamination issue.[1]

Personally, my effort has been to use local models only as of late, and it’s gone pretty well!

[1]: https://z.ai/blog/scaling-pain

1 more reply

j / k navigate · click thread line to collapse