AINeutralarXiv – CS AI · 8h ago6/10
🧠
Bayes-Sufficient Representations in Supervised Learning
A new theoretical framework defines Bayes-sufficient representations in supervised learning, establishing what information is genuinely required for optimal predictions based on loss functions. The work formalizes the concept of Bayes quotients and minimal representations, connecting representation learning to property elicitation theory with experimental validation across synthetic and real datasets.