1 code implementation • 23 Oct 2023 • Jaap Jumelet, Willem Zuidema
With access to the underlying true source, our results show striking differences and outcomes in learning dynamics between different classes of words.
1 code implementation • 19 Oct 2023 • Abhijith Chintam, Rahel Beloch, Willem Zuidema, Michael Hanna, Oskar van der Wal
Language models (LMs) exhibit and amplify many types of undesirable biases learned from the training data, including gender bias.
1 code implementation • 15 Oct 2023 • Hosein Mohebbi, Grzegorz Chrupała, Willem Zuidema, Afra Alishahi
Transformers have become a key architecture in speech processing, but our understanding of how they build up representations of acoustic and linguistic structure is limited.
no code implementations • 5 Oct 2023 • Anna Langedijk, Hosein Mohebbi, Gabriele Sarti, Willem Zuidema, Jaap Jumelet
In recent years, many interpretability methods have been proposed to help interpret the internal states of Transformer-models, at different levels of precision and complexity.
1 code implementation • 21 Jun 2023 • Jaap Jumelet, Willem Zuidema
We study feature interactions in the context of feature attribution methods for post-hoc interpretability.
1 code implementation • 30 Jan 2023 • Hosein Mohebbi, Willem Zuidema, Grzegorz Chrupała, Afra Alishahi
Self-attention weights and their transformed variants have been the main source of information for analyzing token-to-token interactions in Transformer-based models.
no code implementations • 24 Nov 2022 • Oskar van der Wal, Dominik Bachmann, Alina Leidinger, Leendert van Maanen, Willem Zuidema, Katrin Schulz
In particular, we will explore two central notions from psychometrics, the construct validity and the reliability of measurement tools, and discuss how they can be applied in the context of measuring model bias.
1 code implementation • NAACL (GeBNLP) 2022 • Oskar van der Wal, Jaap Jumelet, Katrin Schulz, Willem Zuidema
With full access to the data and to the model parameters as they change during every step while training, we can map in detail how the representation of gender develops, what patterns in the dataset drive this, and how the model's internal state relates to the bias in a downstream task (semantic textual similarity).
1 code implementation • 30 Sep 2021 • Arabella Sinclair, Jaap Jumelet, Willem Zuidema, Raquel Fernández
We investigate the extent to which modern, neural language models are susceptible to structural priming, the phenomenon whereby the structure of a sentence makes the same structure more probable in a follow-up sentence.
no code implementations • COLING 2020 • Phong Le, Willem Zuidema
Interpreting the inner workings of neural models is a key step in ensuring the robustness and trustworthiness of the models, but work on neural network interpretability typically faces a trade-off: either the models are too constrained to be very useful, or the solutions found by the models are too complex to interpret.
1 code implementation • 31 May 2020 • Samira Abnar, Mostafa Dehghani, Willem Zuidema
Having the right inductive biases can be crucial in many tasks or scenarios where data or computing resources are a limiting factor, or where training data is not perfectly representative of the conditions at test time.
7 code implementations • ACL 2020 • Samira Abnar, Willem Zuidema
This makes attention weights unreliable as explanations probes.
1 code implementation • CONLL 2019 • Jaap Jumelet, Willem Zuidema, Dieuwke Hupkes
Extensive research has recently shown that recurrent neural language models are able to process a wide range of grammatical phenomena.
1 code implementation • WS 2019 • Samira Abnar, Lisa Beinborn, Rochelle Choenni, Willem Zuidema
In this paper, we define and apply representational stability analysis (ReStA), an intuitive way of analyzing neural language models.
1 code implementation • 4 Jun 2019 • Samira Abnar, Lisa Beinborn, Rochelle Choenni, Willem Zuidema
In this paper, we define and apply representational stability analysis (ReStA), an intuitive way of analyzing neural language models.
no code implementations • 1 Jun 2019 • Mathijs Mul, Willem Zuidema
We approach this classic question with current methods, and demonstrate that recurrent neural networks can learn to recognize first order logical entailment relations between expressions.
no code implementations • 16 Jan 2019 • Willem Zuidema, Dieuwke Hupkes, Geraint Wiggins, Constance Scharff, Martin Rohrmeier
Human language, music and a variety of animal vocalisations constitute ways of sonic communication that exhibit remarkable structural complexity.
no code implementations • WS 2018 • Mario Giulianelli, Jacqueline Harding, Florian Mohnert, Dieuwke Hupkes, Willem Zuidema
We show that `diagnostic classifiers', trained to predict number from the internal states of a language model, provide a detailed understanding of how, when, and where this information is represented.
1 code implementation • 28 Nov 2017 • Dieuwke Hupkes, Sara Veldhoen, Willem Zuidema
To develop an understanding of what the recurrent network encodes, visualisation techniques alone do not suffice.
no code implementations • WS 2018 • Samira Abnar, Rasyan Ahmed, Max Mijnheer, Willem Zuidema
We evaluate 8 different word embedding models on their usefulness for predicting the neural activation patterns associated with concrete nouns.
no code implementations • WS 2016 • Phong Le, Willem Zuidema
Recursive neural networks (RNN) and their recently proposed extension recursive long short term memory networks (RLSTM) are models that compute representations for sentences, by recursively combining word embeddings according to an externally provided parse tree.
no code implementations • HLT 2015 • Phong Le, Willem Zuidema
We present a self-training approach to unsupervised dependency parsing that reuses existing supervised and unsupervised parsing algorithms.
Ranked #1 on Unsupervised Dependency Parsing on Penn Treebank
1 code implementation • SEMEVAL 2015 • Phong Le, Willem Zuidema
We are proposing an extension of the recursive neural network that makes use of a variant of the long short-term memory architecture.