attribution-decay

MIT CSAIL researchers discovered 'attribution decay' in large generative AI models, where generated images often cannot be traced back to individual training data, using a novel diffusion ensemble architecture.