Proving LLM backdoor-safety without revealing weights: ZKML over formal methods
DimitrisPapail · x · 2026-11-11
A thread argues for a way to prove an LLM satisfies specific backdoor-safety properties without exposing proprietary weights — and to prove such properties succinctly to anyone downstream even for open weights, analogous to not knowing where a Trojan hides. The proposed hammer: ZKML combined with formal methods, offered as a potential defense against newly demonstrated dormant-neuron backdoor attacks.
More from Safety
- Fortinet patches nearly 100 vulnerabilities after adding AI models to security testing — Potential-Couple-745 · 2026-11-11
- New BadEdit backdoors LLMs via model editing with just 15 samples, up to 100% success — terryyuezhuo · 2026-11-11
- Treat weights like code: undetectable malicious payloads in open models spark safety debate — DimitrisPapail · 2026-11-11
- Alice & Bob: ~25,000 cat qubits could factor RSA-2048 in 3 days, slashing prior estimates — AryHHAry · 2026-11-11
- Running local open-source models? Dormant-neuron backdoors mean you may have let the right one in — DimitrisPapail · 2026-11-11
- OpenAI open-sources Codex Security CLI and SDK for AI-driven vulnerability scanning and fixing — openai · 2026-11-11