Today starts the Mathematical Foundations of AI workshop, organised by Flavio and Stefano. The programme brings together diffusion models, transformers, and associative memories, and includes a contributed talk by Chenxiao on our work on heterogeneous data in diffusion models.
Lab updates
News
Talks, papers, grants, workshops, and other moments from the group, ordered by when they happened.
Timeline
The newest updates are first; older lab activity continues down the page.
Great news! Stefano has been awarded 39,000 SEK from Helge Ax:son Johnsons stiftelse, supporting research travel and international scientific exchange.
Stefano is giving a talk on curriculum learning at BrainNet 2026 on May 27.
New preprint from the group! Flavio, Chenxiao, Enrico, Luca, and Stefano released The Interplay of Data Structure and Imbalance in the Learning Dynamics of Diffusion Models, studying how class structure and imbalance shape generalisation and memorisation in diffusion models.
ICML decisions are out, and we are excited to have two accepted papers! Jie, Bruno, and Stefano's work on local minima in high-dimensional two-layer ReLU networks was accepted, and Devon, Richard, Benjamin, Steven, and Stefano's position paper on model collapse and low-resource communities was accepted as a Spotlight.
Great news! Stefano has been awarded a Vetenskapsfond grant from Wilhelm och Martina Lundgrens stiftelser, supporting the group's research on the theoretical foundations of machine learning.
Stefano presented the group's research goals and recent results at the MIND Institute at the University of the Witwatersrand.
Chenxiao and Flavio presented our work on diffusion models and data imbalance at the RAIL Lab and CAANDL, respectively.
Come visit us this June! Flavio and Stefano are organising the workshop Mathematical Foundations of AI in Chalmers, on June 8-10. Many high-profile speakers are joining us, to discuss recent theoretical advancements in Diffusion Models, Transformers and Associative Memories.
Flavio is presenting our ongoing work On class unbalance in diffusion models at the 15th Nordic Workshop on Statistical Physics in NORDITA, as a spotlight talk!
Great news! New positions opening soon. We have been awarded the AI Alignment Project, securing nearly £1M in funding. This will support the hiring of two Postdoctoral Researchers and one Research Assistant to work on developing the theoretical foundations of AI safety.
Stefano will present our work paper on bias dynamics at the MLLS seminars in Copenhagen this week!
EurIPS Workshop decisions are out, and we are excited to see our workshop "Unifying Perspectives on Learning Biases" accepted at the conference! Interested in the topic? Send us your contribution!
We are exited to see the 3rd edition of Analytical Connectionism taking place at UCL (London, UK) with the goal of connecting bias in ML, neuroscience, and psychology! Three members of the group will be there: Chenxiao (as attendee), Flavio (as TA), and Stefano (as organiser).
Flavio and Stefano we will be in Cargese (France) for the Statistical Physics & Machine Learning: moving forward event. Flavio will present the work on the Vector Hopfield model, while Stefano will talk about bias propagation and amplification in AI.
Stefano we will be speaking about the dynamics of bias in the 6th Youth in High-Dimensions Workshop at ICTP (Trieste, Italy).
Next week, on May 26th-27th, we will host the Workshop in Advancements in High-Dimensional Methods for Machine Learning. We are looking forward to the event!
Great news! Stefano will present the work on bias propagation and amplification at the ICLR Workshop on Spurious Correlation and Shortcut Learning!
Save the date! Stefano and Flavio organised two events on statistical physics of learning here in Gothenburg! This is unfortunately a late announcement for the first event (it was on April 1st), but there is still time to join us for our Workshop in Advancements in High-Dimensional Methods for Machine Learning on May 26-27.
We are very happy to meet with the local cognitive science community at the G-Cog Seminar Series. Stefano will talk about our recent work accepted to CogSci conference on curriculum learning and what modern ML theory can tell us about it.
Stefano will talk about our work on learning with prior experience and in particular optimal control in continual learning at the workshop AI for active matter - From plankton to robots.
Great news! Stefano has been nominated Visiting Lecturer at Wits University. While this collaboration with Wits is older, it is nice to see it formally recognised. Stefano and the CAandL Lab, are organising a master-level course on ML theory for year 2026!
It is a great pleasure to meet the statistical physics community of the Nordic countries at the Nordic Workshop on Statistical Physics. Stefano will talk about how StatPhys tools can shed lights on ML problems and in particular about how we can understand bias generation and amplification in ML.
Application for the 3rd edition of Analytical Connectionism! This time we are back to London from Aug 25th to Sep 5th. Application deadline on April 18th.
Great news! We got two papers accepted at ICLR 2025:: 1. On optimal control theory applied to statphys models, and 2. On the impact of initialisation in representation learning.
Stefano will present our NeurIPS paper on bias dynamics at the FAIMI Workshop this week!
Back to Cognitive Science. Stefano will present this aspect of our research at the workshop At the Crossroads of AI and Cognitive Science, here at Chalmers!
We are very excited about this year's Nobel Prizes (and Physics' one in particular)! Stefano and Devdatt Dubhashi will discuss them in an upcoming Colloquium here in Gothenburg. Join us!
In Paris, Stefano will speak at the Informal workshop on the problem of Class Imbalance. Looking forward to discussing ideas and starting new projects!
Stefano will speak at the CHAIR Structured Learning Workshop.
Exciting news! Congratulations to John Hopfield and Geoffrey Hinton for winning the 2024 Nobel Prize! As part of this community, we can’t help but feel a sense of pride and joy. In these two articles, we discuss the achievement with colleagues both at our division and at GU.
Our paper on bias evolution under SGD dynamics (preprint here) has been accepted to NeurIPS 2024! Congrats everyone and especially Anchit Jain for leading the project!
This the start of the group! Stefano in today starting in Chalmers as Assistent Professor!
Today starts the new edition of Analytical Connectionism at the Flatiron Institute in NYC. We have a great set of speakers, TAs and project mentors! Two exciting weeks ahead!
Our paper on transfer learning, probing the dependance of the best source representation on data abundance and similarity, has been accepted to TMLR! Congrats everybody!
Excited to announce two available positions: one for a PhD student and one for a Postdoc. Click the links to learn more and apply!
Delighted that our paper on cognitive fatigue has been accepted to CogSci 2024 as an oral contribution!
Last week to apply for the 2nd edition of Analytical Connectionism! Application deadline on Friday 17th.
We'll open soon two positions for a Postdoc and a PhD student, stay tuned!
ICML decisions are out and we got two papers accepted!
Stefano is teaching a short lecture series on Statistical Physics of Learning at Wits University.
Stefano is co-organising the COSYNE Workshop on The geometry & dynamics of learning: Bridging analytical and experimental insights into neural representations.
Stefano will be visiting Wits University for the next 4 months!
Stefano is visiting the Department of Computing Sciences of Bocconi University thanks to the UK-Italy Trustworthy AI Visiting Researcher Programme.
Stefano is in Santa Barbara at UCSB for the next 6 weeks to follow another great KITP program on Deep Learning from the Perspective of Physics and Neuroscience.
Stefano is co-organising a workshop on Analytical Approaches for Neural Network Dynamics in Paris.