Persona inference from packet-length and inter-arrival-time sequences achieves about 84% accuracy on mixed-site traffic from 10 websites and 15 personas using an LLM-driven browsing framework.
A simple framework for contrastive learning of visual representations
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
baseline 1
citation-polarity summary
years
2026 3verdicts
UNVERDICTED 3roles
baseline 1polarities
baseline 1representative citing papers
Fixed isotropic marginals in JEPAs can be maximally misaligned with unknown structured geometries, and HamJEPA using symplectic Hamiltonian leapfrog maps improves kNN and linear-probe performance on CIFAR-100 and ImageNet-100.
A positive-unlabeled learning method trains a lightweight per-query clone encoder on augmented views of a single anchor to detect media clones in cultural repositories by thresholding latent l2 distances, achieving F1=90.79 on AtticPOT.
citing papers explorer
-
PersonaFingerprint: Measuring Persona Inference on Modern Websites with LLM-Driven Browsing
Persona inference from packet-length and inter-arrival-time sequences achieves about 84% accuracy on mixed-site traffic from 10 websites and 15 personas using an LLM-driven browsing framework.
-
Beyond Isotropy in JEPAs: Hamiltonian Geometry and Symplectic Prediction
Fixed isotropic marginals in JEPAs can be maximally misaligned with unknown structured geometries, and HamJEPA using symplectic Hamiltonian leapfrog maps improves kNN and linear-probe performance on CIFAR-100 and ImageNet-100.
-
Detecting Media Clones in Cultural Repositories Using a Positive Unlabeled Learning Approach
A positive-unlabeled learning method trains a lightweight per-query clone encoder on augmented views of a single anchor to detect media clones in cultural repositories by thresholding latent l2 distances, achieving F1=90.79 on AtticPOT.