• Reference bias in monolingual machine translation evaluation

  Fomicheva, Marina; Specia, Lucia (ACL (Association for Computational Linguistics), 2016)

  In the translation industry, human translations are assessed by comparison with the source texts. In the Machine Translation (MT) research community, however, it is a common practice to perform quality assessment using a ...
• F-to-c-structure mapping: accounting for inflectional morphology and periphrasis

  Alsina i Keith, Àlex (CSLI Publications, 2023)

  The treatment of inflectional periphrasis is problematic in LFG, apparently because of the lexicalist nature of the framework. A close inspection of what is usually understood by lexicalism reveals two distinct, but ...
• Partitivity in romance and the syntax-morphology connection

  Alsina i Keith, Àlex (CSLI Publications, 2022)

  This paper claims that the relationship between morphology and syntax is multidirectional. It argues against the generally accepted position in LFG that word formation feeds the syntax and that syntax cannot feed word ...
• Unnatural language processing: how do language models handle machine-generated prompts?

  Kervadec, Corentin; Franzon, Francesca; Baroni, Marco (ACL (Association for Computational Linguistics), 2023)

  Language model prompt optimization research has shown that semantically and grammatically well-formed manually crafted prompts are routinely outperformed by automatically generated token sequences with no apparent meaning ...
• Bridging information-theoretic and geometric compression in language models

  Cheng, Emily; Kervadec, Corentin; Baroni, Marco (ACL (Association for Computational Linguistics), 2023)

  For a language model (LM) to faithfully model human language, it must compress vast, potentially infinite information into relatively few dimensions. We propose analyzing compression in (pre-trained) LMs from two points ...
• Referential communication in heterogeneous communities of pre-trained visual deep networks

  Mahaut, Matéo; Franzon, Francesca; Dessì, Roberto; Baroni, Marco (International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS), 2023-07-31)

  As large pre-trained image-processing neural networks are being embedded in autonomous agents such as self-driving cars or robots, the question arises of how such systems can communicate with each other about the ...
• Asking between the lines: elicitation of evoked questions in text

  Westera, Matthijs; Rohde, Hannah (Institute for Logic, Language and Computation (ILLC), 2019)

  We introduce a novel, scalable method aimed at annotating potential and actual Questions Under Discussion (QUDs) in naturalistic discourse. It consists of asking naive participants first what questions a certain portion ...
• Run Like a Girl! Sport-Related Gender Bias in Language and Vision

  Harrison, Sophie; Gualdoni, Eleonora; Boleda, Gemma (ACL (Association for Computational Linguistics), 2023)

  Gender bias in Language and Vision datasets and models has the potential to perpetuate harmful stereotypes and discrimination. We analyze gender bias in two Language and Vision datasets. Consistent with prior work, ...
• Neutralizing the effect of translation shifts on automatic machine translation evaluation

  Fomicheva, Marina; Bel Rafecas, Núria; da Cunha Fanego, Iria (Springer, 2015)

  State-of-the-art automatic Machine Translation [MT] evaluation is based on the idea that the closer MT output is to Human Translation [HT], the higher its quality. Thus, automatic evaluation is typically approached by ...
• Quantifying informativeness of names in visual space

  Gualdoni, Eleonora; Kemp, Charles; Xu, Yang; Boleda, Gemma (Cognitive Science Society, 2023)

  The human lexicon expresses a wide array of concepts with a limited set of words. Previous work has suggested that semantic categories are structured compactly to enable informative communication. Informativeness is typically ...
• 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 ...
• 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 ...
• No ordered arguments needed for nouns

  Grimm, Scott; McNally, Louise, 1965- (Institute for Logic, Language and Computation (ILLC), 2013)

  Syntacticians have widely assumed since [11] that there is a fundamental difference between so-called argument structure nominals (AS-nominals, also called Complex Event Nominals), e.g. destruction, and non-AS-nominals, ...
• Does referent predictability affect the choice of referential form? A computational approach using masked coreference resolution

  Aina, Laura; Liao, Xixian; Boleda, Gemma; Westera, Matthijs (ACL (Association for Computational Linguistics), 2021)

  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. Studying these dynamics in the domain of referring expressions has ...
• Challenges in including extra-linguistic context in pre-trained language models

  Sorodoc, Ionut-Teodor; Aina, Laura; Boleda, Gemma (ACL (Association for Computational Linguistics), 2022)

  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). ...
• How universal is metonymy? Results from a large-scale multilingual analysis

  Khishigsuren, Temuulen; Bella, Gábor; Brochhagen, Thomas; Marav, Daariimaa; Giunchiglia, Fausto; Batsuren, Khuyagbaatar (ACL (Association for Computational Linguistics), 2022)
• Introducing QuBERT: a large monolingual corpus and BERT model for Southern Quechua

  Zevallos, Rodolfo; Ortega, John E.; Chen, William; Castro, Richard; Bel Rafecas, Núria; Yoshikawa, Cesar; Ventura, Renzo; Aradiel, Hilario; Melgarejo, Nelsi (ACL (Association for Computational Linguistics), 2022)

  The lack of resources for languages in the Americas has proven to be a problem for the creation of digital systems such as machine translation, search engines, chat bots, and more. The scarceness of digital resources for ...
• Data augmentation for low-resource Quechua ASR improvement

  Zevallos, Rodolfo; Bel Rafecas, Núria; Cámbara Ruiz, Guillermo; Farrús, Mireia; Luque, Jordi (International Speech Communication Association (ISCA), 2022)

  Automatic Speech Recognition (ASR) is a key element in new services that helps users to interact with an automated system. Deep learning methods have made it possible to deploy systems with word error rates below 5% for ...
• Metaphorical measure expressions

  Sutton, Peter R.; Filip, Hana; Snider, Todd; Windhearn, Mia (University of Konstanz, 2021)

  We present an analysis of measure phrases such as heap/ounce of information in which the measure expression receives a non-literal, metaphorical interpretation. We present evidence that these metaphorical measure ...
• Restrictions on copredication: a situation theoretic approach

  Sutton, Peter R. (Cornell University, Linguistic Society of America, 2022)

  This paper proposes a situation theoretic account of polysemy: polysemous nouns denote situations that witness (i.e. contain) multiple entities of different types. For instance, lunch denotes situations that contain an ...

View more