PhD Position: Syntactic and Semantic Invariants for Compositional Semantic Prediction

The Laboratoire d’Informatique de Grenoble (LIG) is seeking a PhD candidate to work on the ANR project 'COMPO: Inductive Biases for Compositionality-capable Deep Learning Models of Natural Language'. The successful applicant will explore inductive biases related to linguistic structures, with the aim of building small NLP models that can infer generalization rules from few data points. The research will focus on developing language representations that are invariant to noise and exploring the use of structural constraints to guide decoding in seq2seq models. The PhD candidate will be supervised by Éric Gaussier and Maximin Coavoux and will be part of the GETALP and APTIKAL teams at LIG.

The starting date is between September and December 2024, and the funding is provided by the ANR project for a duration of 3 years. The salary is approximately 2300€ gross/month. The applicant should have a Master’s degree in NLP, computer science, and experience in NLP and machine learning. To apply, please send a CV, cover letter, and academic transcripts to eric.gaussier at univ-grenoble-alpes.fr and maximin.coavoux at univ-grenoble-alpes.fr.