arXiv:2609.27581v1 Announce Type: new Abstract: Step Law gives power-law formulas for the optimal peak learning rate eta* and batch size B* when pre-training language models. It was calibrated on mod...
The paper introduces a new closed‑form scaling law that extends Chinchilla’s original formula to handle data‑constrained regimes. It decomposes loss into undercapacity, undertraining, and overfitting components, saturating between an irreducible loss and an uninformed baseline. The authors validate the model on diverse architectures and domains, achieving state‑of‑the‑art RMSE across multiple LLM scaling‑law grids and enabling cost‑aware training allocations.
arXiv:2606. 19367v1 Announce Type: new Abstract: Building on a two-parameter Weibull framework for diagnosing transformer weight distributions, we study why the Weibull weight-scale parameter $\lambda$ grows, overshoots, and then relaxes during AdamW training.
arXiv:2607. 21866v1 Announce Type: new Abstract: Prior classical-ML learning-curve work fits power laws to tree, linear, and kernel models on tabular data, but at small scale: typically one curve, one team, a handful of cells.
arXiv:2512. 22088v3 Announce Type: replace-cross Abstract: The scaling law, a cornerstone of Large Language Model (LLM) development, predicts improvements in model performance with increasing computational resources.
arXiv:2607. 23777v1 Announce Type: cross Abstract: The discovery of scaling laws has motivated training neural networks on ever increasing quantities of data.
arXiv:2609.37745v1 Announce Type: new Abstract: Neural scaling, in which loss falls as a power law with training, is central to large language models, and one recent proposal is that a $1/3$ exponent...
arXiv:2607. 11950v1 Announce Type: cross Abstract: Brain field potentials are scale-free: their power spectra follow a $1/f^{\beta}$ law whose aperiodic exponent $\beta$ tracks cortical state, and sleep depth in particular is a shift in $\beta$.
arXiv:2608. 01624v1 Announce Type: cross Abstract: Adapting a language model to a task no longer requires training all of its weights, and a line of parameter-efficient methods has driven the trainable count from billions down to a handful of scalars.
arXiv:2607. 20594v1 Announce Type: cross Abstract: When does a weight-tied looped transformer -- one block applied T times -- implement an actual algorithm?
arXiv:2610.28871v1 Announce Type: new Abstract: Looped transformers reuse one block of layers many times and are often trained by backpropagating through only the last k loops. We measure how much of...
arXiv:2605.08541v3 Announce Type: replace Abstract: Neural scaling laws approximate a language model's loss as a power-law function of parameter count $N$ and token count $D$. Following Chinchilla-st...