arXiv:2606. 16028v1 Announce Type: new Abstract: Modern deep learning architectures are increasingly multi-task and multi-modal, using a pretrained foundation model combined with task-specific, fine-tuned models.
arXiv:2606. 02765v1 Announce Type: cross Abstract: Model dimension ($d_{model}$) is a fundamental hyperparameter in transformer language models, yet its role in setting the geometric limits of feature representation remains under-explored.
arXiv:2606. 02385v1 Announce Type: cross Abstract: Sparse Autoencoders (SAEs) have found success parsing neural representations into interpretable concepts, providing a basis for understanding and control.
arXiv:2512.04696v3 Announce Type: replace Abstract: We develop a flexible feature selection framework based on deep neural networks that approximately controls the false discovery rate (FDR), a measu...
arXiv:2607. 05546v1 Announce Type: cross Abstract: We develop a unified function space theory of deep fully connected neural networks.
arXiv:2606. 09725v1 Announce Type: new Abstract: Disentanglement, the separation of factors of variation in data using neural networks, remains a long-standing challenge in machine learning.
arXiv:2512. 24780v2 Announce Type: replace Abstract: Neural networks trained with standard objectives exhibit behaviors characteristic of probabilistic inference: soft clustering, prototype specialization, and Bayesian uncertainty tracking.
The paper investigates the expressive power of multimodal contrastive learning architectures by treating them as parameterized families of joint density estimators. It shows that the classic two‑tower CLIP model is a universal approximator for two modalities, while a common extension that sums pairwise similarities fails to approximate arbitrary joint distributions when three or more modalities are involved, though it can match all pairwise conditionals. To address this limitation, the authors introduce Hadamard‑CLIP, which adds a single learned weight vector to restore universal approximation for any number of modalities while retaining CLIP’s efficient retrieval capabilities.
arXiv:2605. 22472v2 Announce Type: replace Abstract: Winner-take-all (WTA) networks constitute a central circuit motif in cortical networks of the brain.
arXiv:2606. 19538v1 Announce Type: new Abstract: Convolutional networks, recurrent networks, and transformers each encode different inductive biases -- locality, sequential memory, and content-dependent pairwise interaction -- and have remained mathematically distinct since their inception.
arXiv:2606. 14954v1 Announce Type: cross Abstract: We develop a general framework for analyzing representation costs of parametric data-fitting methods through their parameter-space regularizers.
arXiv:2606. 21497v2 Announce Type: replace-cross Abstract: Modern deep neural networks are trained using error backpropagation, which requires sequential forward and backward computations across network layers.
arXiv:2606. 00130v2 Announce Type: replace-cross Abstract: Large deep neural networks are costly to store and deploy because inference must move and evaluate many parameters.