Feature Lottery? A Bifurcation Theory of Concept Emergence 事件

Feature Lottery? A Bifurcation Theory of Concept Emergence arXiv:2605.24057v1 Announce Type: cross Abstract: Neural networks acquire structured representations at specific moments during training, yet identifying these transitions typically relies on retrospective, label-dependent metrics. We introduce a bifurcation theory of representation dynamics to detect these moments in real time. Analyzing a passive GMM probe attached to the evolving encoder, we show the onset of structure corresponds to