Zobrazeno 1 - 10
of 20 225
pro vyhledávání: '"A, Lampert"'
Autor:
Lampert, Amichai
Let $\mathbb{K}$ be a number field and $f\in \mathbb{K}[x_1,\ldots,x_n]$ a form of degree $d$. Write $s(f)$ for the strength of $f$ and $\bar{s}(f)$ for its strength over $\bar{\mathbb{Q}}.$ We prove that there exists a constant $A = A(d)$ such that
Externí odkaz:
http://arxiv.org/abs/2410.00248
Antennas are more prevalent than ever enabling 5G connectivity for wide ranging applications like cellular communication, IoT, autonomous vehicles, etc. Optimizing an antenna design can be challenging and employing traditional optimization techniques
Externí odkaz:
http://arxiv.org/abs/2408.03674
We study the Kolmogorov-Arnold Network (KAN), recently proposed as an alternative to the classical Multilayer Perceptron (MLP), in the application for differentially private model training. Using the DP-SGD algorithm, we demonstrate that KAN can be m
Externí odkaz:
http://arxiv.org/abs/2407.12569
Autor:
Lampert, Amichai, Snowden, Andrew
Let $K$ be a Birch field, that is, a field for which every diagonal form of odd degree in sufficiently many variables admits a non-zero solution; for example, $K$ could be the field of rational numbers. Let $f_1, \ldots, f_r$ be homogeneous forms of
Externí odkaz:
http://arxiv.org/abs/2406.18498
Autor:
Lampert, Y., Shams-Ansari, A., Gaier, A., Tomasino, A., Rajabali, S., Magalhaes, L., Lončar, M., Benea-Chelmus, I. -C.
Modern communication and sensing technologies rely on connecting the optical domain with the microwave domain. Terahertz technologies, spanning increased frequencies from 100 GHz to 10 THz, are critical for providing larger bandwidths and faster swit
Externí odkaz:
http://arxiv.org/abs/2406.15651
Dynamic link prediction is an important problem considered by many recent works proposing various approaches for learning temporal edge patterns. To assess their efficacy, models are evaluated on publicly available benchmark datasets involving contin
Externí odkaz:
http://arxiv.org/abs/2406.04897
Deep neural networks (DNNs) exhibit a surprising structure in their final layer known as neural collapse (NC), and a growing body of works has currently investigated the propagation of neural collapse to earlier layers of DNNs -- a phenomenon called
Externí odkaz:
http://arxiv.org/abs/2405.14468
Autor:
Kalinin, Nikita, Lampert, Christoph
Current state-of-the-art methods for differentially private model training are based on matrix factorization techniques. However, these methods suffer from high computational overhead because they require numerically solving a demanding optimization
Externí odkaz:
http://arxiv.org/abs/2405.13763
Autor:
Lampert, Amichai
In this note, we present an elementary proof of the fact that the slice rank of a trilinear form over a finite field is bounded above by a linear expression in the analytic rank. The existing proofs by Adiprasito-Kazhdan-Ziegler and Cohen-Moshkovitz
Externí odkaz:
http://arxiv.org/abs/2404.19704
A common way to numerically solve Fokker-Planck equations is the Chang-Cooper method in space combined with one of the Euler methods in time. However, the explicit Euler method is only conditionally positive, leading to severe restrictions on the tim
Externí odkaz:
http://arxiv.org/abs/2404.07641