Výsledky vyhledávání - "Davari, MohammadReza"

Report

Testing No slip model with pulsar timing arrays: NANOGrav and IPTA

Autor: Davari, Mohammadreza, Allahyari, Alireza, Khosravi, Shahram

We perform an observational study of modified gravity considering a potential inflationary interpretation of pulsar timing arrays (PTA). We use a motivated model known as no slip in which the gravitational wave propagation is modified. Specifically,

Externí odkaz: http://arxiv.org/abs/2409.05956

Zobrazit plný text záznamu

Report

Model Breadcrumbs: Scaling Multi-Task Model Merging with Sparse Masks

Autor: Davari, MohammadReza, Belilovsky, Eugene

The rapid development of AI systems has been greatly influenced by the emergence of foundation models. A common approach for targeted problems involves fine-tuning these pre-trained foundation models for specific target tasks, resulting in a rapid sp

Externí odkaz: http://arxiv.org/abs/2312.06795

Zobrazit plný text záznamu

Report

ToxBuster: In-game Chat Toxicity Buster with BERT

Autor: Yang, Zachary, Maricar, Yasmine, Davari, MohammadReza, Grenon-Godbout, Nicolas, Rabbany, Reihaneh

Detecting toxicity in online spaces is challenging and an ever more pressing problem given the increase in social media and gaming consumption. We introduce ToxBuster, a simple and scalable model trained on a relatively large dataset of 194k lines of

Externí odkaz: http://arxiv.org/abs/2305.12542

Zobrazit plný text záznamu

Report

Prototype-Sample Relation Distillation: Towards Replay-Free Continual Learning

Autor: Asadi, Nader, Davari, MohammadReza, Mudur, Sudhir, Aljundi, Rahaf, Belilovsky, Eugene

In Continual learning (CL) balancing effective adaptation while combating catastrophic forgetting is a central challenge. Many of the recent best-performing methods utilize various forms of prior task data, e.g. a replay buffer, to tackle the catastr

Externí odkaz: http://arxiv.org/abs/2303.14771

Zobrazit plný text záznamu

Report

Reliability of CKA as a Similarity Measure in Deep Learning

Autor: Davari, MohammadReza, Horoi, Stefan, Natik, Amine, Lajoie, Guillaume, Wolf, Guy, Belilovsky, Eugene

Comparing learned neural representations in neural networks is a challenging but important problem, which has been approached in different ways. The Centered Kernel Alignment (CKA) similarity metric, particularly its linear variant, has recently beco

Externí odkaz: http://arxiv.org/abs/2210.16156

Zobrazit plný text záznamu

Report

Probing Representation Forgetting in Supervised and Unsupervised Continual Learning

Autor: Davari, MohammadReza, Asadi, Nader, Mudur, Sudhir, Aljundi, Rahaf, Belilovsky, Eugene

Continual Learning research typically focuses on tackling the phenomenon of catastrophic forgetting in neural networks. Catastrophic forgetting is associated with an abrupt loss of knowledge previously learned by a model when the task, or more broadl

Externí odkaz: http://arxiv.org/abs/2203.13381

Zobrazit plný text záznamu

Report

Toponym Identification in Epidemiology Articles - A Deep Learning Approach

Autor: Davari, MohammadReza, Kosseim, Leila, Bui, Tien D.

When analyzing the spread of viruses, epidemiologists often need to identify the location of infected hosts. This information can be found in public databases, such as GenBank, however, information provided in these databases are usually limited to t

Externí odkaz: http://arxiv.org/abs/1904.11018

Zobrazit plný text záznamu