arXiv:2608. 06766v1 Announce Type: cross Abstract: Training changes a network's predictions while allocating task-relevant structure across its internal units.
arXiv:2609.39408v1 Announce Type: cross Abstract: Population loss can remain nearly constant while a neural network learns a substantially more predictive representation. We establish this separation...
The paper derives an exact discrete‑time law that captures how learning‑rate schedules and weight decay interact in scale‑invariant neural networks, showing that a single scalar quantity governs the effective step size. It demonstrates that the balance point between contraction and expansion is intrinsically unstable, leading to recurrent dynamics when using constant learning rates with weight decay. The authors extend this analysis to various optimizers and datasets, confirming the law’s precision and showing that performance peaks sharply at the predicted boundary.
Direct feedback alignment (DFA) trains hidden layers via fixed random projections of output error, but with tanh hidden units and independent sigmoid outputs, plain stochastic gradient descent can stall near a constant predictor of class frequencies. This stall is traced to the error’s common mode—a rank‑one component shared across inputs—that drives tanh units toward saturation. The study shows that calibration of the baseline readout to class priors suppresses collapse and speeds learning, while other interventions such as using Adam, adjusting feedback strength, or subtracting batch means affect the severity and recovery of collapse across MNIST, CIFAR‑10, and deeper networks.
arXiv:2610.01728v1 Announce Type: cross Abstract: Understanding loss landscapes is central to explaining neural-network training, yet their structure remains only partially understood even in simple...
The paper investigates how normalization makes neural networks scale‑invariant, creating a feedback loop between learning‑rate schedules and weight decay that controls the effective step size of the optimizer. It derives an exact discrete‑time law showing that a single scalar quantity captures all schedule and decay effects, with norm growth providing a self‑quenching counter‑force that defines a sharp boundary between contraction‑ and expansion‑dominated regimes. Through exact analysis of a normalized regression model and experiments on MLPs, CNNs, GPT‑2, and various datasets, the authors demonstrate that constant learning rates with weight decay are intrinsically unstable, leading to recurrent dynamics, and that adaptive optimizers exhibit weaker stabilization under normalization. "whyItMatters":"The study provides a precise, actionable rule for controlling training dynamics and schedule design in modern deep learning by isolating a single governing quantity for scale‑invariant optimization."
arXiv:2606. 04476v1 Announce Type: new Abstract: In this paper, we study the gradient descent dynamics for jointly training both layers of a one-hidden-layer ReLU network to fit a linear target function.
arXiv:2606. 05863v1 Announce Type: new Abstract: Grokking suggests that fitting the training data and learning a simple underlying rule may occur on different time scales.
arXiv:2607. 03613v1 Announce Type: new Abstract: We study the implicit bias of noisy stochastic gradient descent in training wide two-layer ReLU networks for multivariate regression.
arXiv:2607. 10869v1 Announce Type: new Abstract: We study the population gradient flow of an infinitely wide two-layer neural network learning a misspecified single-index model in high dimension.
The paper investigates how two‑layer polynomial‑width neural networks learn orthogonal multi‑index targets under standard initialization. It shows that incremental learning still occurs: the loss decreases sequentially following the Hermite expansion, with lower‑order components learned first. The dynamics also exhibit a competitive reallocation of parameter mass, shifting into the target subspace and concentrating on aligned neurons. The analysis uses a symmetry‑based finite‑width approximation and demonstrates that vanilla gradient descent displays the same qualitative behavior.
arXiv:2605. 15435v2 Announce Type: replace Abstract: Standard deep-learning pipelines usually choose the network architecture before training and keep it fixed throughout optimization.
Neural Cellular Automata (NCAs) are shown to learn general, scale‑invariant topological primitives in their hidden channels, which can be transferred from a teacher to a student model for few‑shot learning. The study introduces a transfer‑learning mechanism that injects pretrained hidden states into a student, improving early optimization and outperforming recurrent and feed‑forward baselines on MNIST benchmarks with only ~9,800 parameters. Mechanistic analysis reveals that hidden channels decouple feature extraction from classification, converging to mutually orthogonal states that absorb morphological complexity.