[mattdesl]

The full title of the paper is “Questioning Representational Optimism in Deep Learning: The Fractured Entangled Representation Hypothesis.”

https://arxiv.org/abs/2505.11581