4 ms·
I’m skeptical of these side channel attacks that rely on training a neural network on specific controlled scenarios on controlled hardware. I believe that with
by Aurornis 5mo ago
I’m skeptical of these side channel attacks that rely on training a neural network on specific controlled scenarios on controlled hardware. I believe that with enough time and effort and the perfect circumstances where the user is only visiting their website and doing one other thing that the network was trained on it can match.
It does not seem useful as a general purpose side channel vector.
- hansvm 5mo agoIt depends what you mean by "general purpose." First, these things generalize more often than you'd expect. Second, even in the absence of generalization they're still useful for, e.g., fingerprinting activities to manufacture a unique ID where non previously existed.
- Aurornis 5mo agoThe paper isn’t describing a unique ID fingerprint. It’s looking for specific activity patterns to match against training data of running specific commands on specific hardware.
- ramenat2am 5mo agoThat's basically just a research, theoretical attack vector. It doesn't mean it's viable for general purpose old school mass privacy invasion
- 1970-01-01 5mo agoPublish or perish. It worked once, in a controlled lab, mostly (80-90% guess). Good enough for more millions in funding.. Not really joking here. https://hannesweissteiner.com/ https://hannesweissteiner.com/ https://hannesweissteiner.com/publications/frost/ https://hannesweissteiner.com/publications/frost/