Antoine Simoulin
Graduate student @ University of Paris, Paris, France
transformers
depth
dynamic
concreteness
abstractness
visual representations
3
presentations
SHORT BIO
I am currently completing my Ph.D. in Natural Language Processing at Paris University in a joint program sponsored by Quantmetry. I graduated from ENSTA Paris in mathematical optimization and obtained a Master (M2) from Ecole Polytechnique in Data Science. My research focuses on building standalone sentence embeddings. I examine sentence encoder architectures, as well as training and evaluation methods. I contribute to the open-source community with the creation of datasets and implementation of deep learning models. I specifically contribute to resources for French.
Presentations
Unifying Parsing and Tree-Structured Models for Generating Sentence Semantic Representations
Antoine Simoulin and 1 other author
How Many Layers and Why? An Analysis of the Model Depth in Transformers
Antoine Simoulin and 1 other author