AI RESEARCH
Visual Disentangled Diffusion Autoencoders: Scalable Counterfactual Generation for Foundation Models
arXiv CS.LG
•
ArXi:2601.21851v2 Announce Type: replace Foundation models, despite their robust zero-shot capabilities, remain vulnerable to spurious correlations and 'Clever Hans' strategies. Existing mitigation methods often rely on unavailable group labels or computationally expensive gradient-based adversarial optimization. To address these limitations, we propose Visual Disentangled Diffusion Autoencoders (DiDAE), a novel framework integrating frozen foundation models with disentangled dictionary learning for efficient, gradient-free counterfactual generation directly for the foundation model.