57 citations · 61 across the 7 of their papers we have counts for
7 papers · 1 filter
Enhancing Interpretability using Human Similarity Judgements to Prune Word Embeddings
Natalia Flechas Manrique, Wanqian Bao, Aurelie Herbelot +1
Interpretability methods in NLP aim to provide insights into the semantics underlying specific system architectures. Focusing on word embeddings, we present a supervised-learning m…
CALaMo: a Constructionist Assessment of Language Models
Ludovica Pannitto, Aurélie Herbelot
This paper presents a novel framework for evaluating Neural Language Models' linguistic abilities using a constructionist approach. Not only is the usage-based model in line with t…
Novel Aficionados and Doppelgängers: a referential task for semantic representations of individual entities
Andrea Bruera, Aurélie Herbelot
In human semantic cognition, proper names (names which refer to individual entities) are harder to learn and retrieve than common nouns. This seems to be the case for machine learn…
Recurrent babbling: evaluating the acquisition of grammar from limited input data
Ludovica Pannitto, Aurélie Herbelot
Recurrent Neural Networks (RNNs) have been shown to capture various aspects of syntax from raw linguistic input. In most previous experiments, however, learning happens over unreal…
How to marry a star: probabilistic constraints for meaning in context
Katrin Erk, Aurelie Herbelot
In this paper, we derive a notion of 'word meaning in context' that characterizes meaning as both intensional and conceptual. We introduce a framework for specifying local as well…
High-risk learning: acquiring new word vectors from tiny data
Aurelie Herbelot, Marco Baroni
Distributional semantics models are known to struggle with small data. It is generally accepted that in order to learn 'a good vector' for a word, a model must have sufficient exam…
Pay Attention to Those Sets! Learning Quantification from Images
Ionut Sorodoc, Sandro Pezzelle, Aurélie Herbelot +2
Major advances have recently been made in merging language and vision representations. But most tasks considered so far have confined themselves to the processing of objects and le…