Browsing by Author "Baroni, Marco"

Sort by: Order: Results:

  • Can discrete information extraction prompts generalize across language models?
    Rakotonirina, Nathanael Carraz; Dessì, Roberto; Petroni, Fabio; Riedel, Sebastian; Baroni, Marco (International Conference on Learning Representations (ICLR), 2023)
    We study whether automatically-induced prompts that effectively extract information from a language model can also be used, out-of-the-box, to probe other language models for the same information. After confirming that ...
  • CNNs found to jump around more skillfully than RNNs: compositional generalization in seq2seq convolutional networks
    Dessi, Roberto; Baroni, Marco (ACL (Association for Computational Linguistics), 2019)
    Lake and Baroni (2018) introduced the SCAN dataset probing the ability of seq2seq models to capture compositional generalizations, such as inferring the meaning of “jump around” 0- shot from the component words. ...
  • Communicating artificial neural networks develop efficient color-naming systems
    Chaabouni, Rahma; Kharitonov, Eugene; Dupoux, Emmanuel; Baroni, Marco (National Academy of Sciences, 2021)
    Words categorize the semantic fields they refer to in ways that maximize communication accuracy while minimizing complexity. Focusing on the well-studied color domain, we show that artificial neural networks trained with ...
  • Communication breakdown: on the low mutual intelligibility between human and neural captioning
    Dessì, Roberto; Gualdoni, Eleonora; Franzon, Francesca; Boleda, Gemma; Baroni, Marco (ACL (Association for Computational Linguistics), 2022)
    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. We conduct this comparison on the recently ...
  • Controlled tasks for model analysis: retrieving discrete information from sequences
    Sorodoc, Ionut-Teodor; Boleda, Gemma; Baroni, Marco (ACL (Association for Computational Linguistics), 2021)
    In recent years, the NLP community has shown increasing interest in analysing how deep learning models work. Given that large models trained on complex tasks are difficult to inspect, some of this work has focused ...
  • Cross-domain image captioning with discriminative finetuning
    Dessì, Roberto; Bevilacqua, Michele; Gualdoni, Eleonora; Rakotonirina, Nathanael Carraz; Franzon, Francesca; Baroni, Marco (Institute of Electrical and Electronics Engineers (IEEE), 2023)
    Neural captioners are typically trained to mimic humangenerated references without optimizing for any specific communication goal, leading to problems such as the generation of vague captions. In this paper, we show ...
  • Distributional memory: a general framework for corpus-based semantics
    Baroni, Marco; Lenci, Alessandro (MIT Press, 2010)
    Research into corpus-based semantics has focused on the development of ad hoc models that treat single tasks, or sets of closely related tasks, as unrelated challenges to be tackled by extracting different kinds of ...
  • Distributional vectors encode referential attributes
    Boleda, Gemma; Gupta, Abhijeet; Baroni, Marco; Padó, Sebastian (ACL (Association for Computational Linguistics), 2015)
    Distributional methods have proven to excel at capturing fuzzy, graded aspects of meaning (Italy is more similar to Spain than to Germany). In contrast, it is difficult to extract the values of more specific attributes of ...
  • EEG responds to conceptual stimuli and corpus semantics
    Baroni, Marco; Lenci, Alessandro (ACL (Association for Computational Linguistics), 2009)
    Mitchell et al. (2008) demonstrated that corpus-extracted models of semantic knowledge can predict neural activation patterns recorded using fMRI. This could be a very powerful technique for evaluating conceptual models ...
  • Emergent language generalization and acquisition speed are not tied to compositionality
    Kharitonov, Eugene; Baroni, Marco (ACL (Association for Computational Linguistics), 2020)
    Studies of discrete languages emerging when neural agents communicate to solve a joint task often look for evidence of compositional structure. This stems for the expectation that such a structure would allow languages ...
  • High-risk learning: acquiring new word vectors from tiny data
    Herbelot, Aurélie; Baroni, Marco (ACL (Association for Computational Linguistics), 2018)
    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 examples of its usage. This contradicts the ...
  • Hubness and pollution: delving into cross-space mapping for zero-shot learning
    Lazaridou, Angeliki; Dinu, Georgiana; Baroni, Marco (ACL (Association for Computational Linguistics), 2015)
    Zero-shot methods in language, vision and other domains rely on a cross-space mapping function that projects vectors from the relevant feature space (e.g., visualfeature-based image representations) to a large semantic ...
  • Intensionality was only alleged: on adjective-noun composition in distributional semantics
    Boleda, Gemma; Baroni, Marco; Pham, Nghia The; McNally, Louise, 1965- (ACL (Association for Computational Linguistics), 2013)
    Distributional semantics has very successfully modeled semantic phenomena at the word level, and recently interest has grown in extending it to capture the meaning of phrases via semantic composition. We present experiments ...
  • Interpretable agent communication from scratch (with a generic visual processor emerging on the side)
    Dessì, Roberto; Kharitonov, Eugene; Baroni, Marco (Neural Information Processing Systems (NeurIPS), 2021)
    As deep networks begin to be deployed as autonomous agents, the issue of how they can communicate with each other becomes important. Here, we train two deep nets from scratch to perform large-scale referent identification ...
  • Jointly optimizing word representations for lexical and sentential tasks with the C-PHRASE model
    Pham, Nghia The; Kruszewski, German; Lazaridou, Angeliki; Baroni, Marco (ACL (Association for Computational Linguistics), 2015)
    We introduce C-PHRASE, a distributional semantic model that learns word representations by optimizing context prediction for phrases at all levels in a syntactic tree, from single words to full sentences. C-PHRASE outperforms ...
  • Large linguistically-processed web corpora for multiple languages
    Baroni, Marco; Kilgarriff, Adam (ACL (Association for Computational Linguistics), 2006)
    The Web contains vast amounts of linguistic data. One key issue for linguists and language technologists is how to access it. Commercial search engines give highly compromised access. An alternative is to crawl the Web ...
  • Leveraging preposition ambiguity to assess compositional distributional models of semantics
    Ritter, Samuel; Long, Cotie; Paperno, Denis; Baroni, Marco; Botvinick, Matthew; Goldberg, Adele (ACL (Association for Computational Linguistics), 2015)
    Complex interactions among the meanings of words are important factors in the function that maps word meanings to phrase meanings. Recently, compositional distributional semantics models (CDSM) have been designed with the ...
  • Linguistic generalization and compositionality in modern artificial neural networks.
    Baroni, Marco (Royal Society, 2020)
    In the last decade, deep artificial neural networks have achieved astounding performance in many natural language processing tasks. Given the high productivity of language, these models must possess e ective generalization ...
  • Living a discrete life in a continuous world: reference with distributed representations
    Boleda, Gemma; Padó, Sebastian; Pham, Nghia The; Baroni, Marco (ACL (Association for Computational Linguistics), 2017)
    Reference is the crucial property of language that allows us to connect linguistic expressions to the world. Modeling it requires handling both continuous and discrete aspects of meaning. Data-driven models excel at the ...
  • Mechanisms for handling nested dependencies in neural-network language models and humans
    Lakretz, Yair; Hupkes, Dieuwke; Vergallito, Alessandra; Marelli, Marco; Baroni, Marco; Dehaene, Stanislas (Elsevier, 2021)
    Recursive processing in sentence comprehension is considered a hallmark of human linguistic abilities. However, its underlying neural mechanisms remain largely unknown. We studied whether a modern artificial neural network ...