Zobrazeno 1 - 10
of 15 068
pro vyhledávání: '"A., Mehri"'
Why do gradient-based explanations struggle with Transformers, and how can we improve them? We identify gradient flow imbalances in Transformers that violate FullGrad-completeness, a critical property for attribution faithfulness that CNNs naturally
Externí odkaz:
http://arxiv.org/abs/2411.16760
Atrial fibrillation (AF), the most common cardiac arrhythmia, is associated with heart failure and stroke. Accurate segmentation of the left atrium (LA) in 3D late gadolinium-enhanced (LGE) MRI is helpful for evaluating AF, as fibrotic remodeling in
Externí odkaz:
http://arxiv.org/abs/2411.05963
Autor:
Mehri, Zohra, Boudjemaa, Abdelaali
We investigate analytically and numerically the Anderson localization of quasiparticles in binary Bose mixtures in the presence of the Lee-Huang-Yang quantum and thermal corrections subjected to correlated disordered potentials. We calculate the dens
Externí odkaz:
http://arxiv.org/abs/2410.17884
Large Language Models (LLMs) have revolutionized natural language understanding and generation tasks but suffer from high memory consumption and slow inference times due to their large parameter sizes. Traditional model compression techniques, such a
Externí odkaz:
http://arxiv.org/abs/2410.09615
Quantifying fiber disarray, which is a prominent maladaptation associated with hypertrophic cardiomyopathy, remains critical to understanding the disease's complex pathophysiology. This study investigates the role of heterogeneous impairment of fiber
Externí odkaz:
http://arxiv.org/abs/2409.15508
Modern power systems increasingly demand converter-driven generation systems that integrate seamlessly with grid infrastructure. Grid-based converters are particularly advantageous, as they operate in harmony with conventional synchronous machines. H
Externí odkaz:
http://arxiv.org/abs/2409.11548
Autor:
D'Oosterlinck, Karel, Xu, Winnie, Develder, Chris, Demeester, Thomas, Singh, Amanpreet, Potts, Christopher, Kiela, Douwe, Mehri, Shikib
Large Language Models (LLMs) are often aligned using contrastive alignment objectives and preference pair datasets. The interaction between model, paired data, and objective makes alignment a complicated procedure, sometimes producing subpar results.
Externí odkaz:
http://arxiv.org/abs/2408.06266
This paper investigates how group-control can be effectively used for motion planning for microrobot swarms in a global field. We prove that Small-Time Local Controllability (STLC) in robot positions is achievable through group-control, with the mini
Externí odkaz:
http://arxiv.org/abs/2406.13829
We propose SLoPe, a Double-Pruned Sparse Plus Lazy Low-rank Adapter Pretraining method for LLMs that improves the accuracy of sparse LLMs while accelerating their pretraining and inference and reducing their memory footprint. Sparse pretraining of LL
Externí odkaz:
http://arxiv.org/abs/2405.16325
The advent of 5G technology promises a paradigm shift in the realm of telecommunications, offering unprecedented speeds and connectivity. However, the efficient management of traffic in 5G networks remains a critical challenge. It is due to the dynam
Externí odkaz:
http://arxiv.org/abs/2405.05239