A pose-conditioned large-margin contrastive encoder isolates persistent biometric identity cues from transmitted latents in talking-head videoconferencing to flag impersonation attacks via cosine similarity without inspecting the output video.
arXiv preprint arXiv:2301.12345 , year =
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
representative citing papers
Diverse ensembles of prompted and fine-tuned GPT-4.1-Mini monitors achieve 2.4x better detection of flawed code solutions than homogeneous ensembles on adversarial inputs.
A single photon sequentially probing five emulated absorbers on a programmable photonic processor can reveal all five without being absorbed, matching theoretical efficiency predictions.
citing papers explorer
-
Unmasking Puppeteers: Leveraging Biometric Leakage to Expose Impersonation in AI-Based Videoconferencing
A pose-conditioned large-margin contrastive encoder isolates persistent biometric identity cues from transmitted latents in talking-head videoconferencing to flag impersonation attacks via cosine similarity without inspecting the output video.
-
Ensemble Monitoring for AI Control: Diverse Signals Outweigh More Compute
Diverse ensembles of prompted and fine-tuned GPT-4.1-Mini monitors achieve 2.4x better detection of flawed code solutions than homogeneous ensembles on adversarial inputs.
-
Interaction-free measurement of multiple objects using a universal integrated photonic processor
A single photon sequentially probing five emulated absorbers on a programmable photonic processor can reveal all five without being absorbed, matching theoretical efficiency predictions.