no code implementations • 4 May 2023 • Ghazi Felhi
The second model, QKVAE, uses separate latent variables to form keys and values for its Transformer decoder and is able to separate syntactic and semantic information in its neural representations.
1 code implementation • 22 Jun 2022 • Ghazi Felhi, Joseph Le Roux, Djamé Seddah
Starting from a deep probabilistic generative model with attention, we measure the interaction between latent variables and realizations of syntactic roles and show that it is possible to obtain, without supervision, representations of sentences where different syntactic roles correspond to clearly identified different latent variables.
1 code implementation • NAACL 2022 • Ghazi Felhi, Joseph Le Roux, Djamé Seddah
In the attention of Transformers, keys handle information selection while values specify what information is conveyed.
1 code implementation • EMNLP (insights) 2021 • Ghazi Felhi, Joseph Le Roux, Djamé Seddah
We compare the simplified versions to standard SSVAEs on 4 text classification tasks.
1 code implementation • 24 Dec 2020 • Ghazi Felhi, Joseph Le Roux, Djamé Seddah
We present an unsupervised method to obtain disentangled representations of sentences that single out semantic content.
2 code implementations • 13 Oct 2020 • Ghazi Felhi, Joseph Leroux, Djamé Seddah
Even though Variational Autoencoders (VAEs) are widely used for semi-supervised learning, the reason why they work remains unclear.
no code implementations • JEPTALNRECITAL 2020 • Davide Buscaldi, Ghazi Felhi, Dhaou Ghoul, Joseph Le Roux, Ga{\"e}l Lejeune, Xu-Dong Zhang
Dans notre travail nous nous sommes int{\'e}ress{\'e} {\`a} deux questions : celle du choix de la mesure du similarit{\'e} d{'}une part et celle du choix des op{\'e}randes sur lesquelles se porte la mesure de similarit{\'e}.
no code implementations • 19 Sep 2019 • Léo Hemamou, Ghazi Felhi, Jean-Claude Martin, Chloé Clavel
In this paper, we focus on studying influential non verbal social signals in asynchronous job video interviews that are discovered by deep learning methods.
no code implementations • 25 Jul 2019 • Léo Hemamou, Ghazi Felhi, Vincent Vandenbussche, Jean-Claude Martin, Chloé Clavel
As part of a project to help recruiters, we collected a corpus of more than 7000 candidates having asynchronous video job interviews for real positions and recording videos of themselves answering a set of questions.