Andrea Gregor de Varda
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Andrea Gregor de Varda has a total of 19 co-authors. Here are the top co-authors by number of shared publications:
- Evelina Fedorenko
McGovern Institute for Brain Research, Massachusetts Institute of Technology, 02139
6 shared publications - Andrew Kyle Lampinen
Google DeepMind
4 shared publications - Ferdinando Pio D’ELia
Center for Language Technology, University of Copenhagen
4 shared publications - Hope Kean
Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology
4 shared publications Marco Marelli
3 shared publications- Marco Marelli
Department of Psychology, University of Milano-Bicocca, Milano, Italy
3 shared publications
Latest Publications
- Open Access
Behavioral and brain responses to language reflect different levels of linguistic representation
Human language processing can be studied through both behavior and brain activity, yet it remains unclear whether these two data types reflect sensitivity to the same information. One influential view holds that both behavioral and neural responses are largely determined by processing effort, often estimated by word surprisal together with the context-independent properties of word frequency and length. At the same time, neural responses have been shown to encode richer aspects of linguistic content, including meaning. Here, we use neural network language models to operationalize these alternatives and systematically compare, within the same analytic computational framework, the predictive power of low-dimensional effort-based predictors and high-dimensional embedding representations that encode contextualized linguistic content, including meaning. Across 8 behavioral datasets and 5 neural datasets (4 fMRI and 1 ERP), we find that processing effort captures substantial variance in both behavioral and neural measures of language processing, in line with much previous work. However, for brain responses---but not for behavioral measures---embedding representations carry substantial predictive power beyond the estimates of processing effort. These results therefore suggest that neural data provide access to rich, high-dimensional dynamics of language comprehension, whereas behavioral data reflect a bottlenecking of these dynamics into a small set of theoretically motivated properties of contextualized linguistic input.
Open AccessBehavioral and brain responses to language reflect different levels of linguistic representation
Human language processing can be studied through both behavior and brain activity, yet it remains unclear whether these two data types reflect sensitivity to the same information. One influential view holds that both behavioral and neural responses are largely determined by processing effort, often estimated by word surprisal together with the context-independent properties of word frequency and length. At the same time, neural responses have been shown to encode richer aspects of linguistic content, including meaning. Here, we use neural network language models to operationalize these alternatives and systematically compare, within the same analytic computational framework, the predictive power of low-dimensional effort-based predictors and high-dimensional embedding representations that encode contextualized linguistic content, including meaning. Across 8 behavioral datasets and 5 neural datasets (4 fMRI and 1 ERP), we find that processing effort captures substantial variance in both behavioral and neural measures of language processing, in line with much previous work. However, for brain responses---but not for behavioral measures---embedding representations carry substantial predictive power beyond the estimates of processing effort. These results therefore suggest that neural data provide access to rich, high-dimensional dynamics of language comprehension, whereas behavioral data reflect a bottlenecking of these dynamics into a small set of theoretically motivated properties of contextualized linguistic input.
- Open Access
- 10 April 2026
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 123 (17)
Reply to Hu: Reasoning traces need not be faithful to account for the cost of human thinking
Reply to Hu: Reasoning traces need not be faithful to account for the cost of human thinking
- 10 April 2026
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 123 (17)
Open AccessReply to Hu: Reasoning traces need not be faithful to account for the cost of human thinking
Reply to Hu: Reasoning traces need not be faithful to account for the cost of human thinking
- Open Access
- 16 March 2026
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 123 (12)
Reply to Vankov et al.: Reasoning traces are linked to accuracy and capture key dimensions of problem complexity
- 16 March 2026
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 123 (12)
Open AccessReply to Vankov et al.: Reasoning traces are linked to accuracy and capture key dimensions of problem complexity
- Open Access
- 20 January 2026
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 123 (4)
Reply to Dujmović: The alignment in cost between human and model reasoning is an empirical phenomenon worth explaining
- 20 January 2026
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 123 (4)
Open AccessReply to Dujmović: The alignment in cost between human and model reasoning is an empirical phenomenon worth explaining
- Open Access
- 8 December 2025
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 20 (12) , e0337947
IconicITA: Iconicity ratings of the Italian affective lexicon
Iconicity, defined as the potential of linguistic signs to resemble properties or features of their referents, is increasingly recognized as a general property of language. One common approach for quantifying iconicity is to collect iconicity ratings. Although iconicity datasets have been developed for several languages, no comprehensive dataset of iconicity ratings is currently available for Italian. The current study presents IconicITA , the first dataset of Italian iconicity ratings for the 1,121 words of the Italian adaptation of Affective Norms for English Words (ANEW). Ratings were collected from both Italian native speakers (L1) and English native speakers with Italian as a second language (L2). Including L2 participants allowed us to contribute to the debate on whether iconicity ratings genuinely measure form-meaning resemblance, rather than exclusively reflecting semantic properties. We showed that L1 Italian iconicity ratings are positively associated with perceptual strength in the auditory and haptic modalities, and with specificity ratings. Conversely, we found a negative correlation between iconicity and concreteness, age of acquisition, word frequency, and letter frequency. In general, the relationship between Italian iconicity norms and various psycholinguistic variables largely replicated previous findings in the literature on iconicity. Considering L2 data, the ratings provided by L2 speakers correlated more strongly with the Italian L1 data compared to the translation-equivalent English L1 data. This finding suggests that participants’ judgments were influenced not only by the semantic information of the words but also by language-specific form-level properties. We take this result as evidence of the validity of iconicity ratings to operationalize the degree of resemblance between words’ form and meaning.
- 8 December 2025
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 20 (12) , e0337947
Open AccessIconicITA: Iconicity ratings of the Italian affective lexicon
Iconicity, defined as the potential of linguistic signs to resemble properties or features of their referents, is increasingly recognized as a general property of language. One common approach for quantifying iconicity is to collect iconicity ratings. Although iconicity datasets have been developed for several languages, no comprehensive dataset of iconicity ratings is currently available for Italian. The current study presents IconicITA , the first dataset of Italian iconicity ratings for the 1,121 words of the Italian adaptation of Affective Norms for English Words (ANEW). Ratings were collected from both Italian native speakers (L1) and English native speakers with Italian as a second language (L2). Including L2 participants allowed us to contribute to the debate on whether iconicity ratings genuinely measure form-meaning resemblance, rather than exclusively reflecting semantic properties. We showed that L1 Italian iconicity ratings are positively associated with perceptual strength in the auditory and haptic modalities, and with specificity ratings. Conversely, we found a negative correlation between iconicity and concreteness, age of acquisition, word frequency, and letter frequency. In general, the relationship between Italian iconicity norms and various psycholinguistic variables largely replicated previous findings in the literature on iconicity. Considering L2 data, the ratings provided by L2 speakers correlated more strongly with the Italian L1 data compared to the translation-equivalent English L1 data. This finding suggests that participants’ judgments were influenced not only by the semantic information of the words but also by language-specific form-level properties. We take this result as evidence of the validity of iconicity ratings to operationalize the degree of resemblance between words’ form and meaning.
- Open Access
- 19 November 2025
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 122 (47)
The cost of thinking is similar between large reasoning models and humans
Do neural network models capture the cognitive demands of human reasoning? Across seven reasoning tasks, we show that the length of the chain-of-thought generated by large reasoning models predicts human reaction times both within tasks—tracking item-level difficulty—and across tasks—capturing broader differences in cognitive demands. This model-to-human alignment shows that out-of-the-box reasoning models reflect core features underlying problem and task complexity in human cognition, without requiring any built-in symbolic mechanisms.
- 19 November 2025
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 122 (47)
Open AccessThe cost of thinking is similar between large reasoning models and humans
Do neural network models capture the cognitive demands of human reasoning? Across seven reasoning tasks, we show that the length of the chain-of-thought generated by large reasoning models predicts human reaction times both within tasks—tracking item-level difficulty—and across tasks—capturing broader differences in cognitive demands. This model-to-human alignment shows that out-of-the-box reasoning models reflect core features underlying problem and task complexity in human cognition, without requiring any built-in symbolic mechanisms.
- 1 September 2025
- Published by Springer Nature in Nature Reviews Psychology
- Vol. 4 (10) , 623
Learning meaning from latent patterns in language use
- 1 September 2025
- Published by Springer Nature in Nature Reviews Psychology
- Vol. 4 (10) , 623
Learning meaning from latent patterns in language use
- 4 June 2025
- Published by Springer Nature in Nature Human Behaviour
- Vol. 9 (7) , 1339-1341
High variability in LLMs’ analogical reasoning
- 4 June 2025
- Published by Springer Nature in Nature Human Behaviour
- Vol. 9 (7) , 1339-1341
High variability in LLMs’ analogical reasoning
- Open Access
- 22 February 2025
- Published by Elsevier in Journal of Memory and Language
- Vol. 142
A distributional model of concepts grounded in the spatial organization of objects
- 22 February 2025
- Published by Elsevier in Journal of Memory and Language
- Vol. 142
Open AccessA distributional model of concepts grounded in the spatial organization of objects
- Open Access
Multilingual Computational Models Capture a Shared Meaning Component in Brain Responses across 21 Languages
At the heart of language neuroscience lies a fundamental question: How does the brain process the rich variety of languages? Multilingual neural network models offer a way to answer this question by representing linguistic content across languages in a shared space. Leveraging these advances, we evaluated the similarity of linguistic representations in speakers of 21 languages. We combined existing (12 languages across 4 language families) and newly collected fMRI data (9 languages across 4 families) to test encoding models predicting brain activity in the language network using representations from multilingual models. Model representations reliably predicted brain responses within each language. Critically, encoding models can be transferred zero-shot across languages, so that a model trained to predict brain activity in a set of languages can account for responses in a held-out language. These results imply a shared cross-lingual component, which appears to be related to a shared meaning space.
Open AccessMultilingual Computational Models Capture a Shared Meaning Component in Brain Responses across 21 Languages
At the heart of language neuroscience lies a fundamental question: How does the brain process the rich variety of languages? Multilingual neural network models offer a way to answer this question by representing linguistic content across languages in a shared space. Leveraging these advances, we evaluated the similarity of linguistic representations in speakers of 21 languages. We combined existing (12 languages across 4 language families) and newly collected fMRI data (9 languages across 4 families) to test encoding models predicting brain activity in the language network using representations from multilingual models. Model representations reliably predicted brain responses within each language. Critically, encoding models can be transferred zero-shot across languages, so that a model trained to predict brain activity in a set of languages can account for responses in a held-out language. These results imply a shared cross-lingual component, which appears to be related to a shared meaning space.