Upcoming InDeep Masterclass:
Interpreting and Understanding LLMs and Other Deep Learning Models
You are warmly invited to our next InDeep Masterclass Interpreting and Understanding LLMs and other Deep Learning Models, scheduled for Thursday, 4 December 2025. Registration is open!
This masterclass will feature a series of insightful presentations and a hands-on tutorial focused on explainability techniques for Large Language Models (LLMs) and other deep learning architectures. Participants will gain both insights and practical experience in interpreting and understanding the inner workings of modern AI systems. We look forward to welcoming you to this engaging and informative event.

Event details
Date: Thursday 4 December 2025
Time: 09.00-12.30
Location: Deloitte Amsterdam
Participation is free of charge, but places are limited.
Speakers:
– Antske Fokkens (VU)
Presentation on explanatory evaluation.
– Gabriele Sarti (RUG Groningen)
Presentation and hands-on tutorial on Interpreting Large Language Models
– Jelle Zuidema (UvA) “Too Much Information: How do we generate comprehensible and faithful explanations when the input or output is… just a bit much?” Discussing RAG attribution methods, ‘reasoning traces’, and surrogate models.
Upcoming Workshop:
How Computing is Changing the World

Friday 31 October
Workshop “How Computing is Changing the World: Exploring Synergies and Challenges of AI and Quantum Technologies for Society”
The QISS and InDeep research groups are pleased to invite you to a workshop exploring the synergies and challenges of AI and Quantum Technologies for society.
Novel computing technologies, including artificial intelligence and quantum computing, are rapidly transforming society. In this workshop, we will take stock of the recent advances in both fields, examine their ethical and legal dimensions and oversight, and discuss how to shape their societal impacts responsibly. To what extent do advances in these field reinforce one another? What lessons can successful interventions in one field offer the other? Join us for an engaging exchange at the intersection of technology, ethics, and society.

Event details
Date: Friday 31 October 2025
Time: 09.30-18.00
Location: Sustainalab Matrix One, Science Park 301, Amsterdam
Workshop includes lunch, with drinks at the end of the workshop.
Participation is free of charge, but places are limited.
Speakers
• Amira Abbas (Google Quantum AI) TBC
• Mehrnoosh Sadrzadeh (UCL London)
• Evert van Nieuwenburg (Leiden University)
• Ronald de Wolf (CWI & UvA, QuSoft)
• Joris van Hoboken (UvA)
• Martha Lewis (UvA)
• Sebastian De Haro (UvA)
• Jelle Zuidema (UvA)
• Christian Schaffner (UvA, QuSoft)
• Eline de Jong (UvA)
• Matteo Fabbri (UvA)
More speakers will followHow Computing is Changing the World: Exploring Synergies and Challenges of AI and Quantum Technologies for Society
AI is a burgeoning field, with many tangible societal implications, including in our daily lives and for the global economy. Likewise, quantum technologies of various sorts are pervasive, and second-generation quantum technologies such as quantum computing are expected to follow AI’s lead in the future. Furthermore, recent research is moving towards integrating aspects of human and artificial intelligence with quantum technology.
Next to the similarities and connections between the two fields, there are also important differences. For example, quantum computers are in their early stages of development, and their societal impact is yet to be realized. According to the Collingridge dilemma, in the early stages of a technology, it is very difficult to predict its impact while it is still relatively easy to control and steer. Once the technology is fully developed and its impacts are clear, it becomes difficult to control or change its course. Are AI and quantum at different stages of a Collingridge dilemma, so that the impacts of AI are clear but difficult to control, while the impacts of quantum technologies are still relatively easy to control but difficult to predict? What are the interesting cross-connections between these two fields, so that research on the technical and the ethical, legal, and societal (ELS) aspects of these technologies can mutually inform and develop fruitful synergies? Is the framing in terms of two stages of a Collingridge dilemma oversimplified in that, like quantum, AI also faces many uncertainties and future unpredictabilities? And, given that state players and companies are already acting in order to protect themselves against the threat of quantum computers, what methodologies are suitable for anticipating ELS impacts and relevant aspects of quantum? What is the status of the integration of human and artificial intelligence with quantum technology, and what are its possible scientific and societal implications?
This workshop aims to bring together experts from the fields of AI and quantum to explore these questions, foster interdisciplinary dialogue, and identify potential pathways for responsible development and integration of AI and quantum technologies.
Programme 09.30-18.00:
• 09.30 – 10.00 Coffee & arrival
• Presentations and panel discussion: Technologies
10.00 – 10.30 Jelle Zuidema (UvA)
10.30 – 11.00 Mehrnoosh Sadrzadeh (UCL London): ‘From Many Body Quantum Physics to Affordable Vision-language Models via Tensor Networks’
11.00 – 11.15 short break
11.15 – 11.45 Martha Lewis (UvA)
11.45 – 12.15 Ronald de Wolf (CWI & UvA, QuSoft): ‘The Current State of Quantum Computing’
12.15 – 12.45 Panel Discussion, also with Christian Schaffner (UvA, QuSoft), chair Joris van Hoboken (UvA)
• 12:45 – 13.45 Lunch break (same building)
• Presentations and panel discussion: Society
13.45 – 14.00 Introduction, Sebastian De Haro (UvA)
14.00 – 14.30 Evert van Nieuwenburg (Leiden University)
14.30 – 15.00 Eline de Jong (UvA): ‘AI and Quantum Through a Lens of Continuity’
15.00 – 15.15 short break
15.15 – 15.45 TBA
15.45 – 16.15 Matteo Fabbri (UvA)
16.15 – 17.00 Panel Discussion, chair Jelle Zuidema (UvA)
• 17.00 – 18.00 Drinks
QISS and InDeep are research projects part of the National Research Agenda, working together for the first time to bring you this timely workshop.
– The Quantum Impact on Societal Security (QISS) consortium is based at the University of Amsterdam and analyses the ethical, legal and societal impact of the upcoming society-wide transition to quantum-safe cryptography. The consortium’s objective is to contribute to the creation of a Dutch ecosystem where quantum-safe cryptography can thrive, and mobilize this ecosystem to align technological applications with ethical, legal, and social values.
– InDeep is a 6 year research consortium focused on Interpreting Deep Learning Models for Text and Sound, in which 5 universities and 7 companies collaborate. InDeep’s goal is to find ways to explain how Artificial Intelligence processes and generates language, speech and music.
This workshop is made possible with funding from the Quantum Delta Netherlands growthfund program.
InDeep at Interspeech 2025

Many InDeep researchers presented at the Interspeech 2025 conference in Rotterdam, one of the largest and most prominent conferences on speech and speech technology world-wide. Therefore, we also organized a social event including drinks and snacks.
This was InDeep at Interspeech 2025:
Sunday 17 August
– Speech Science Festival 10.00-17.30. InDeep provided a demonstration of emotion-decoding and removal. There were many other demonstrations for a broad audience.
– Tutorial, 15.30-18.30: InDeep members organized the tutorial Interpretability Techniques for Speech Model.
– InDeep Social, 19.00: InDeep held a social event with drinks and snacks in the Stuurboord bar at the Foodhallen.
Monday 18 August
– Special Session on Interpretability in Audio and Speech Technology, 11:00-13:00. Session with three papers by InDeep authors:
+ On the reliability of feature attribution methods for speech classification.
+ Word stress in self-supervised speech models: A cross-linguistic comparison.
+ What do self-supervised speech models know about Dutch? Analyzing advantages of language-specific pre-training.
Consortium Meeting Groningen

On Friday 20 September 2024 InDeep met in the Groningen House of Connections for the next Consortium Meeting. In the morning RUG researchers Daniel Herrman (Philosophy of AI) and Yevgen Matusevych (Cognitive Plausibility of Modern LMs) gave illuminating talks on standards for belief representations in LLMs and bias in visually grounded speech models.

In the afternoon highlights from the various InDeep academic partners were shared, there was a poster session by researchers on their latest research results, as well as fruitful discussions on the future of interpretability methods in language, speech and music models and future directions of the InDeep project.

Consortium Meeting Amsterdam

On Friday 8 March 2024 InDeep had a Consortium meeting in the Amsterdam University Library. It was a day with both talks on recent progress in research, a tutorial, and lots of opportunity to discuss the strengths and weaknesses of interpretability techniques in text, translation, speech and music.
InDeep Masterclass: Explaining Foundation Models


On Thursday 2 November 2023, InDeep organized a Masterclass on Explaining Foundation Models. The event featured some inspiring talks, hands-on experience with explainability techniques for Large Language Models, and the opportunity to share insights and experiences. After the Masterclass there was a Meetup organized by Indeep and Amsterdam Al, on Alternatives for ChatGPT: How good are open source and in-house LLMs? To learn more about this event, you can download here an overview of the presentations and slides.