AI Models Can Provably Hide Arbitrary Capabilities · HackerPapers discussions

AI Models Can Provably Hide Arbitrary Capabilities — mentioned by angry_octet

Tue, 03 Nov 2026 22:12:38 GMT

Embedded models and triggering: https://arxiv.org/abs/2610.13928 You can add layers to an existing model without massive training costs, which is of course obvious, but ideally you would adjust the weights of a number of layers. Ideally this would be done without knowing the weight decryption inputs, i.e. just a worm behaviour. I recommend the tool hashdeep to compute and verify weight checksums trees, and signatures of those checksums with gpg or minisign [1], so you can always verify weights before they are loaded into your engine. https://github.com/jedisct1/minisign

Paper and discussion history. A mention is not an endorsement.

AI Models Can Provably Hide Arbitrary Capabilities — mentioned by solyin

Tue, 03 Nov 2026 17:15:42 GMT

Yes, company A paid company B to say that company A is safe. Company B is under no duress at all with their findings being what company paid them to find. "Just trust these bros that I've recently paid" You didn't address my question on reflections on trusting trust. For the AI era, there's https://arxiv.org/abs/2610.13928

Paper and discussion history. A mention is not an endorsement.

AI Models Can Provably Hide Arbitrary Capabilities — mentioned by AlexCoventry

Tue, 20 Oct 2026 04:15:10 GMT

AI Models Can Provably Hide Arbitrary Capabilities https://arxiv.org/abs/2610.13928 > Capability evaluations aim to surface dangerous model behaviors before deployment, but their reliability depends on the assumption that hidden capabilities can be elicited. We show this assumption does not hold under adversarial conditions by constructing backdoor attacks that embed encrypted neural networks within host model weights, decrypting and executing them only upon receiving a secret trigger. Using digital lockers, these hidden circuits remain provably hard to elicit and interpret under…

Paper and discussion history. A mention is not an endorsement.

AI Models Can Provably Hide Arbitrary Capabilities — HN submission

Tue, 13 Oct 2026 02:07:08 GMT

AI models can provably hide arbitrary capabilities

282 points; 145 discussion replies on HN. Paper and source history. A submission is not an endorsement.