1 code implementation • EMNLP (BlackboxNLP) 2021 • Ionut-Teodor Sorodoc, Gemma Boleda, Marco Baroni
In recent years, the NLP community has shown increasing interest in analysing how deep learning models work.
no code implementations • insights (ACL) 2022 • Ionut Sorodoc, Laura Aina, Gemma Boleda
To successfully account for language, computational models need to take into account both the linguistic context (the content of the utterances) and the extra-linguistic context (for instance, the participants in a dialogue).
no code implementations • 10 Oct 2024 • Eleonora Gualdoni, Gemma Boleda
Human lexicons contain many different words that speakers can use to refer to the same object, e. g., "purple" or "magenta" for the same shade of color.
1 code implementation • 16 Nov 2023 • Yunke He, Xixian Liao, Jialing Liang, Gemma Boleda
Different speakers often produce different names for the same object or entity (e. g., "woman" vs. "tourist" for a female tourist).
no code implementations • 23 May 2023 • Sophia Harrison, Eleonora Gualdoni, Gemma Boleda
A computational model trained on these naming data reproduces the bias.
1 code implementation • 20 Oct 2022 • Roberto Dessì, Eleonora Gualdoni, Francesca Franzon, Gemma Boleda, Marco Baroni
We compare the 0-shot performance of a neural caption-based image retriever when given as input either human-produced captions or captions generated by a neural captioner.
1 code implementation • CoNLL (EMNLP) 2021 • Laura Aina, Xixian Liao, Gemma Boleda, Matthijs Westera
It is often posited that more predictable parts of a speaker's meaning tend to be made less explicit, for instance using shorter, less informative words.
1 code implementation • COLING 2020 • Carina Silberer, Sina Zarrie{\ss}, Matthijs Westera, Gemma Boleda
We also find that standard evaluations underestimate the actual effectiveness of the naming model: on the single-label names of the original dataset (Visual Genome), it obtains −27{\%} accuracy points than on MN v2, that includes all valid object names.
no code implementations • ACL 2020 • Ionut-Teodor Sorodoc, Kristina Gulordava, Gemma Boleda
Language models keep track of complex information about the preceding context {--} including, e. g., syntactic relations in a sentence.
no code implementations • LREC 2020 • Carina Silberer, Sina Zarrie{\ss}, Gemma Boleda
We highlight the challenges involved and provide a preliminary analysis of the ManyNames data, showing that there is a high level of agreement in naming, on average.
no code implementations • 8 Apr 2020 • Kristina Gulordava, Thomas Brochhagen, Gemma Boleda
We find that constraints in both learning and selection can foster mutual exclusivity, as long as they put words in competition for lexical meaning.
no code implementations • 5 Nov 2019 • Alba Herrera-Palacio, Carles Ventura, Carina Silberer, Ionut-Teodor Sorodoc, Gemma Boleda, Xavier Giro-i-Nieto
The goal of this work is to segment the objects in an image that are referred to by a sequence of linguistic descriptions (referring expressions).
1 code implementation • ACL 2019 • Laura Aina, Kristina Gulordava, Gemma Boleda
In neural network models of language, words are commonly represented using context-invariant representations (word embeddings) which are then put in context in the hidden layers.
no code implementations • WS 2019 • Matthijs Westera, Gemma Boleda
Our proposal sheds light on the role of distributional semantics in a broader theory of language and cognition, its relationship to formal semantics, and its place in computational models.
1 code implementation • NAACL 2019 • Laura Aina, Carina Silberer, Matthijs Westera, Ionut-Teodor Sorodoc, Gemma Boleda
In this paper we analyze the behavior of two recently proposed entity-centric models in a referential task, Entity Linking in Multi-party Dialogue (SemEval 2018 Task 4).
no code implementations • 6 May 2019 • Gemma Boleda
Distributional semantics provides multi-dimensional, graded, empirically induced word representations that successfully capture many aspects of meaning in natural languages, as shown in a large body of work in computational linguistics; yet, its impact in theoretical linguistics has so far been limited.
no code implementations • EMNLP 2018 • Kristina Gulordava, Laura Aina, Gemma Boleda
Recent state-of-the-art neural language models share the representations of words given by the input and output mappings.
1 code implementation • NAACL 2019 • Marco Del Tredici, Raquel Fernández, Gemma Boleda
We present the first exploration of meaning shift over short periods of time in online communities using distributional representations.
no code implementations • 5 Aug 2018 • Abhijeet Gupta, Gemma Boleda, Sebastian Pado
Our paper closes this gap by investigating and modeling the lexical relation of instantiation, which holds between an entity-denoting and a category-denoting expression (Marie Curie -- scientist or Mumbai -- city).
1 code implementation • SEMEVAL 2018 • Laura Aina, Carina Silberer, Ionut-Teodor Sorodoc, Matthijs Westera, Gemma Boleda
This paper describes our winning contribution to SemEval 2018 Task 4: Character Identification on Multiparty Dialogues.
no code implementations • WS 2017 • Gemma Boleda
We use language to talk about the world, and so reference is a crucial property of language.
no code implementations • SEMEVAL 2017 • Abhijeet Gupta, Gemma Boleda, Sebastian Pad{\'o}
Word embeddings are supposed to provide easy access to semantic relations such as {``}male of{''} (man{--}woman).
no code implementations • EACL 2017 • Gemma Boleda, Abhijeet Gupta, Sebastian Pad{\'o}
Instances ({``}Mozart{''}) are ontologically distinct from concepts or classes ({``}composer{''}).
no code implementations • 6 Feb 2017 • Gemma Boleda, Sebastian Padó, Nghia The Pham, Marco Baroni
Reference is a crucial property of language that allows us to connect linguistic expressions to the world.
no code implementations • 28 Jun 2016 • Gemma Boleda, Sebastian Padó, Marco Baroni
One of the most basic functions of language is to refer to objects in a shared scene.
3 code implementations • ACL 2016 • Denis Paperno, Germán Kruszewski, Angeliki Lazaridou, Quan Ngoc Pham, Raffaella Bernardi, Sandro Pezzelle, Marco Baroni, Gemma Boleda, Raquel Fernández
We introduce LAMBADA, a dataset to evaluate the capabilities of computational models for text understanding by means of a word prediction task.
no code implementations • 31 Jul 2014 • Alvaro Corral, Gemma Boleda, Ramon Ferrer-i-Cancho
In all cases Zipf's law is fulfilled, in the sense that a power-law distribution of word or lemma frequencies is valid for several orders of magnitude.