Recent & Upcoming Talks

Natural languages have so-called typological universals. Can recent advancements in neural language modeling and cognitive modeling provide insights …

Entity tracking in language models Keeping track of how states and relations of entities change as a text or dialog unfolds is a key prerequisite to …

Methods in natural language processing (NLP) have matured to the point where they can address complex real-world problems. However, the process of …

We present a setup for training, evaluating and interpreting neural language models, that uses artificial, language-like data. The data is generated …

Classical probability distributions on sets of sequences can be modeled using quantum states. Here, we do so with a quantum state that is pure and …

Generative AI has matured to a point where large-scale models can generate text that seems indistinguishable from human-written text and remarkably …

The continued growth of LLMs and their wide-scale adoption in commercial applications such as chatGPT make it increasingly important to (a) develop …

For years, the progress in modeling has outpaced the evaluation in NLP, where we relied predominantly on string-based matching metrics. In this talk, …

Any unique language production context affords speakers with multiple plausible communicative intents, and any intent can be produced in multiple …

Children start to communicate and use language in social interactions from a very young age. This allows them to experiment with their developing …

One of the more surprising (or at least, counter-intuitive) findings in current sentence processing is the fact that fully ambiguous structures like …

Construction Grammar (CxG) is a paradigm from cognitive linguistics that emphasises the connection between syntax and semantics. Rather than rules …

AI tools are ubiquitous, but most users treat it as a black box: a handy tool that suggests purchases, flags spam, or autocompletes text. While …

I present three case studies in which formal languages offer natural notions of generalization in neural networks.

Explainability in question answering allows researchers to check that the model is making the right decision for the right reason. Datasets of …

Systematic generalization is one of the most important open problems of neural networks: given a model trained to solve a certain problem, it will often fail on a test problem with different statistics than the training one, even if the problem should be solvable by the same algorithm.

This talk provides an introduction to text-editing models and a closer look at two models: LaserTagger and EdiT5.

While advances in automated fact-checking are critical in the fight against the spread of misinformation in social media, we argue that more attention …

In this talk we will look at Combinatory Categorial Grammar (CCG) on three distinct, complementary levels of analysis.

With pre-trained transformer-based models continuously increasing in size, there is a dire need for parameter-efficient and modular transfer learning strategies. In this talk, we will touch base on adapter-based fine-tuning, where instead of fine-tuning all weights of a model, small neural network components are introduced at every layer.

Neural networks and other machine learning models compute continuous representations, while humans communicate mostly through discrete symbols. …

I will present two advances in significantly improving the speed of dependency parsing (EMNLP, 2021) and considerably improving the accuracy of unsupervised constituency parsing (Findings of ACL, 2022).

Overview of a couple of projects that involved transformer encoder and decoder recombination, analysis and exploitation. This includes modular and …

Recent work on large language models relies on the intuition that most natural language processing tasks can be described via natural language instructions and that models trained on these instructions show strong zero-shot performance on several standard datasets. However, these models, even though impressive, still perform poorly on a wide range of tasks outside of their respective training and evaluation sets.

There is wide debate about the degree to which the properties of human cognition affect how languages are structured and how they change over time. This controversy extends to the lexicon. We show that the decline of lexical items can be partially accounted for by biases that have been demonstrated in the cognitive science literature.

Our world is extremely complex, and yet we are able to exchange our thoughts and beliefs about it using a relatively small number of words. What computational principles can explain this extraordinary ability?

Recent mobile app technology lets people systematize the process of messaging their friends to urge them to vote. Prior to the most recent US midterm …

Translating text to events happening in a 3D virtual world is one way of evaluating machine understanding of human language. This entails several challenges including how to represent sentences and full discourses that are grounded in the physical and social world.

Although the success of transformers is widely acknowledged in the problem of modelling word sequences, they are in fact general-purpose learners, capable of modelling virtually any kind of sequential data. In this talk, I dispute whether a general-purpose learner like this is well-suited to the task of language learning

Transformers - the purely attention based NN architecture - have emerged as a powerful tool in sequence processing. But how does a transformer think?

This talk presents work in designing systems that use knowledge graphs as structural scaffolds for commonsense reasoning in QA systems.

This talk delves into some of the challenges underlying reasoning – making natural language understanding decisions that depend on multiple, interdependent, models, and exemplify it using the domain of Reasoning about Time, as it is expressed in natural language.