LeakyLMs: Stealing Architecture and Inference Optimizations via Timing
niloofar_mire · x · 2026-08-27
Research titled 'LeakyLMs' demonstrates a side-channel attack stealing model architecture and inference optimizations via per-token timing. The attack reveals hidden draft models and context windows (e.g., a 128K context draft in Gemini Flash 2.5) by observing latency spikes during speculative decoding. Additionally, by deriving runtime-scaling terms from the Transformer graph, the method can recover architectural parameters like hidden size and layer count from black-box endpoints, validated against Llama 3.1 8B.
Related event: Researchers steal hidden LLM architecture via streaming API timing(3 posts)→
More from Safety
- CognitionX Dubai demo walks through full Microsoft Entra ID kill chain, from discovery to Global Admin — lee_stott · 2026-10-15
- DiPLab report: Scale AI's Outlier runs 16 trackers harvesting biometric data from EU workers — AntonioCasilli · 2026-10-15
- Your breathing gives you away: Wi-Fi signals can reveal your breath rhythm — TinfoilTricorn · 2026-10-15
- Cohere acquires Aisle, whose defensive cyber tech found 200+ High/Critical CVEs in curl, FFmpeg and more — aidangomez · 2026-10-15
- Podcast: AI is making pentesting more profitable and bug bounty more competitive — rez0__ · 2026-10-15
- Chatbots' bad health advice leaves no trace, warn researchers in Nature Health paper — alex_verem · 2026-10-15