Adaptive Hierarchical Prior Alignment for Diffusion Transformers
Discover AHPA, a dynamic multi-level alignment method that boosts diffusion transformers' training efficiency and output quality without extra inference co...
Cryptographic Defense Against Dependency Confusion Attacks
Discover how cryptographic registry provenance secures AI package ecosystems against dependency confusion attacks with dual-signature and namespace binding...
SHIELD Dataset & Models for Clinical Note De-identification
Discover SHIELD, a diverse clinical note dataset and distilled language models for efficient, enterprise-scale de-identification of sensitive health data.
Spectral Structure & Equivalence in Multilabel Fisher Discriminants
Explore the spectral structure and objective equivalence of orthogonal multilabel Fisher discriminants with new algebraic and statistical insights.
Copula Correction for Robust Treatment Effect Estimation
Discover a copula-based method to correct endogeneity in doubly robust treatment effect estimation, improving accuracy in healthcare research findings.