CP-MLPs: A Tensor-Rank Theory of Tied and Untied MLP Blocks
arXiv:2611.13647v1 Announce Type: cross Abstract: Transformer feed-forward blocks combine several design choices: squared activations, gating, and tied versus u…
言栖资讯收录 251313 篇 AI、开发与建站文章,按发布时间排序,支持搜索与按语言筛选。
arXiv:2611.13647v1 Announce Type: cross Abstract: Transformer feed-forward blocks combine several design choices: squared activations, gating, and tied versus u…
arXiv:2611.13623v1 Announce Type: cross Abstract: Under i.i.d. uniform sampling from a closed Riemannian manifold embedded in Euclidean space, we derive first-o…
arXiv:2611.13565v1 Announce Type: cross Abstract: Scientific design often requires balancing competing objectives that are expensive to evaluate through experim…
arXiv:2611.13545v1 Announce Type: cross Abstract: BF16 is widely used to pretrain large language models because it has the same dynamic range as FP32 and conver…
arXiv:2611.13460v1 Announce Type: cross Abstract: We study the regression of a scalar response $Y$ on a Euclidean predictor $\mathbf{X} \in \mathbb{R}^p$ and a …
arXiv:2611.13383v1 Announce Type: cross Abstract: We study the excess risk of constant-stepsize online stochastic gradient descent (SGD) with tail-averaging for…
arXiv:2611.13212v1 Announce Type: cross Abstract: GARCH-LSTM hybrid models have become increasingly popular in the volatility forecasting literature. To date, t…
arXiv:2611.12955v1 Announce Type: cross Abstract: Time-series classifiers deployed in risk-sensitive domains must be (1) accurate, (2) interpretable, and (3) ab…
arXiv:2611.12908v1 Announce Type: cross Abstract: Seismic performance assessment relies on computationally expensive nonlinear simulations, limiting uncertainty…
arXiv:2611.12107v1 Announce Type: cross Abstract: Flow-based and diffusion-based models have become the principal means of learning and sampling high-dimensiona…
arXiv:2611.14463v1 Announce Type: new Abstract: Cross-validation estimates predictive performance by repeatedly splitting the data into training and validation …
arXiv:2611.14275v1 Announce Type: new Abstract: Compositional data have been analysed in the literature from three main points of view: directly on the simplex,…
arXiv:2611.13973v1 Announce Type: new Abstract: We consider the problem of estimating the power spectral density of a Vector-valued Autoregressive Moving-Averag…
arXiv:2611.13869v1 Announce Type: new Abstract: We study how outliers affect the hard-margin linear support vector machine when the dimension grows while the sa…
arXiv:2611.13489v1 Announce Type: new Abstract: Simulation optimization typically returns a single estimated optimal point. When decisions also depend on criter…
arXiv:2611.13485v1 Announce Type: new Abstract: Reliable inference for rankings of large language models based on human pairwise comparisons must preserve the p…
arXiv:2611.13337v1 Announce Type: new Abstract: We compute the VC-dimension of metric ball classifiers in the model spaces of Riemannian geometry (Euclidean, sp…
arXiv:2611.13295v1 Announce Type: new Abstract: This paper proposes a novel general Gaussian process framework for change point detection in time series. A chan…
arXiv:2611.13228v1 Announce Type: new Abstract: Multi-horizon forecasting has been dominated by discrete forecast strategies that treat model outputs as fixed-v…
arXiv:2611.13219v1 Announce Type: new Abstract: Causal mediation analysis decomposes a treatment effect into indirect pathways through mediators and direct path…