Mattia Opper

representation learning (3) self-supervised learning (2) self-structuring autoencoder (2) compositional generalization (1) parameter efficient (1) semantic embedding (1) inductive bia (1) text embedding (1) autoencoder architecture (1) word embedding (1) multilingual embedding (1) embedding quality (1) large language model (1) neural network (1) hierarchical composition (1) symbolic structure (1) tree-structured representation (1) neurosymbolic computation (1) contrastive learning (1) cross-lingual transfer (1)