Zobrazeno 1 - 10
of 439 049
pro vyhledávání: '"A Roman"'
Author profiling is the task of inferring characteristics about individuals by analyzing content they share. Supervised machine learning still dominates automatic systems that perform this task, despite the popularity of prompting large language mode
Externí odkaz:
http://arxiv.org/abs/2409.04122
Autor:
Xue, Tengfei, Li, Xuefeng, Azim, Tahir, Smirnov, Roman, Yu, Jianhui, Sadrieh, Arash, Pahlavan, Babak
Large language models (LLMs) have significantly improved code generation, particularly in one-pass code generation. However, most existing approaches focus solely on generating code in a single programming language, overlooking the potential of lever
Externí odkaz:
http://arxiv.org/abs/2409.04114
Here we introduce the application of Tensor Networks (TN) to launch attacks on symmetric-key cryptography. Our approaches make use of Matrix Product States (MPS) as well as our recently-introduced Flexible-PEPS Quantum Circuit Simulator (FQCS). We co
Externí odkaz:
http://arxiv.org/abs/2409.04125
Autor:
Schnabel, Roman, Korobko, Mikhail
GEO 600, Kagra, LIGO, and Virgo were built to observe gravitational waves at frequencies in the audio band, where the highest event rates combined with the largest signal to noise ratios had been predicted. Currently, hypothetical sources of cosmolog
Externí odkaz:
http://arxiv.org/abs/2409.03019
The advancements in audio generative models have opened up new challenges in their responsible disclosure and the detection of their misuse. In response, we introduce a method to watermark latent generative models by a specific watermarking of their
Externí odkaz:
http://arxiv.org/abs/2409.02915
Autor:
Wasil, Akash R., Barnett, Peter, Gerovitch, Michael, Hauksson, Roman, Reed, Tom, Miller, Jack William
International AI governance agreements and institutions may play an important role in reducing global security risks from advanced AI. To inform the design of such agreements and institutions, we conducted case studies of historical and contemporary
Externí odkaz:
http://arxiv.org/abs/2409.02779
Autor:
Shvydkoy, Roman, Teolis, Trevor
The $\mathrm{s}$-model, introduced in arXiv:2211.00117, is an alignment model with the property that the strength of the alignment force, $\mathrm{s}$, is transported along an averaged velocity field. Inspired by the 1D threshold regularity criterion
Externí odkaz:
http://arxiv.org/abs/2409.02409
Autor:
Sarapin, Roman
Consider the $n\times n$ matrix $X_n=A_n+H_n$, where $A_n$ is a $n\times n$ matrix (either deterministic or random) and $H_n$ is a $n\times n$ matrix independent from $A_n$ drawn from complex Ginibre ensemble. We study the limiting eigenvalue distrib
Externí odkaz:
http://arxiv.org/abs/2409.02314
Autor:
Herráiz-López, Víctor, Roca-Jerat, Sebastián, Gallego, Manuel, Ferrández, Ramón, Carrete, Jesús, Zueco, David, Román-Roche, Juan
We study the ground-state phase diagram of an unfrustrated antiferromagnetic Ising chain with longitudinal and transverse fields in the full range of interactions: from all-to-all to nearest-neighbors. First, we solve the model analytically in the st
Externí odkaz:
http://arxiv.org/abs/2409.02165
While differentially private synthetic data generation has been explored extensively in the literature, how to update this data in the future if the underlying private data changes is much less understood. We propose an algorithmic framework for stre
Externí odkaz:
http://arxiv.org/abs/2409.00322