Irene Cannistraci

Postdoctoral Researcher and Lecturer @ ETH Zurich

Postdoctoral Researcher and Lecturer at ETH Zurich in the Institute of Machine Learning, working on representation learning, multimodal machine learning, foundation models, efficiency and post-training. I received my PhD with honors in 2025, and one of my works has been recognized with an ICLR 2024 Spotlight (top 5%).

I am also an ELLIS member, co-organizer of the UniReps workshop series at NeurIPS and the ELLISxUniReps speaker series, and the co-leader of the Computer Science Network of Women at ETH (CSNOW).

Feel free to reach out at irene.cannistraci[at]inf.ethz.ch!

Google Scholar X LinkedIn GitHub CV
Irene Cannistraci

News 🎉

Apr 2026

Tabular Foundation Models Are Effectively Shallow has been accepted to the SD4H and FMSD ICML 2026 Workshops in Seoul!🇰🇷

Apr 2026

TOAST: Transformer Optimization using Adaptive and Simple Transformations has been accepted AS-IS in TMLR!

Mar 2026

2 papers accepted to TMLR! Structure is Supervision: Multiview Masked Autoencoders for Radiology and You Only Train Once: Differentiable Subset Selection for Omics Data has been accepted in Transactions on Machine Learning Research (TMLR).

Mar 2026

3 papers accepted at ICLR 2026 Workshops in Rio🏄 with an oral contribution! Rethinking Machine Unlearning (TTU, Oral), From Leads to Latents (GRaM), and Beyond Independent Frames (FM4Science).

Dec 2025

Co-Organizer and Panel Moderator at UniReps NeurIPS 2025! Panelsession with Sara Hooker, Ahmad Beirami, and Meenakshi Khosla.

Experience

Postdoctoral Researcher, ETH Zurich

Feb 2025 - Present
Institute of Machine Learning, ETH Zurich, MDS Lab · Zurich, Switzerland ongoing

Leading research on representation learning, multimodal ML, foundation models, efficiency and post-training. Supervising MSc and PhD students. Part of the MDS Lab, led by Prof. Julia Vogt.

Representation LearningMultimodal LearningPost-TrainingEfficiencyFoundation Models

Lecturer, ETH Zurich

Feb 2026 - May 2026
Computer Science Department, ETH Zurich · Zurich, Switzerland

Lecturer for the Spring Semester MSc course "Machine Learning for Health Care" (150+ students), co-taught with Prof. G. Rätsch and Prof. V. Boeva. Designed and delivered lectures on interpretability, explainability, fairness, and language models.

HealthcareInterpretabilityFairnessLanguage Models

International Research Visit, Helmholtz AI

Feb 2024 - Jul 2024
Helmholtz AI Munich, AIDOS Lab · Munich, Germany

Research visit working on representation learning and geometric deep learning in the AIDOS Lab, led by Prof. Bastian Rieck.

Representation LearningGeometric Deep Learning

Ph.D. in Computer Science — with Honors

Nov 2020 - Jan 2025
Sapienza University of Rome, GLADIA Lab · Rome, Italy

Thesis: "Improving Neural Networks Efficiency via Representation Similarities". Supervised by Prof. Emanuele Rodolà.

Neural NetworksRepresentation LearningEfficiencyFoundation models

Selected Publications

Full list on Google Scholar →

Tabular Foundation Models Are Effectively Shallow

Irene Cannistraci, Julia Vogt

ICML 2026, FMSD Workshop · 2026 · link

TL;DR: We show that up to 94% of blocks in tabular transformers can be replaced with a closed-form linear translator while largely preserving downstream performance.

TOAST: Transformer Optimization using Adaptive and Simple Transformations

Irene Cannistraci, Simone Antonelli, Emanuele Palumbo, Thomas Sutter, Emanuele Rodolà, Bastian Rieck, Julia Vogt

TMLR · 2026 · link

TL;DR: We propose a training-free method to replace redundant blocks in vision transformers with simple closed-form linear transformations with little to no loss in downstream accuracy.

Rethinking Machine Unlearning: Models Designed to Forget via Key Deletion Under Review

Sonia Laguna, Jorge da Silva Gonçalves, Moritz Vandenhirtz, Alain Ryser, Julia E Vogt*, Irene Cannistraci*

Under Review (*Equal senior authorship) · 2026 · link

TL;DR: We introduce MUNKEY, a machine unlearning method for vision transformers that externalizes instance-specific memorization. Here, unlearning results in a zero-shot key deletion operation.

From Bricks to Bridges: Product of Invariances to Enhance Latent Space Communication Spotlight · Top 5%

Irene Cannistraci, Luca Moschella, Marco Fumero, Valentino Maiorca, Emanuele Rodolà

ICLR 2024 · 2024 · link

TL;DR: We propose a method to incorporate invariances into neural representations to construct a product space of invariant components and unlock applications, such as merging, stitching, and reusing different neural modules.

Bootstrapping Parallel Anchors for Relative Representations

Irene Cannistraci, Luca Moschella, Valentino Maiorca, Marco Fumero, Antonio Norelli, Emanuele Rodolà

ICLR 2023, Tiny Papers · 2023 · link

TL;DR: We propose an optimization-based method to expand limited semantic correspondences between domains and enable latent space communication.

From Charts to Atlas: Merging Latent Spaces into One

Donato Crisostomi, Irene Cannistraci, Luca Moschella, Pietro Barbiero, Marco Ciccone, Pietro Liò, Emanuele Rodolà

NeurIPS 2023, NeurReps Workshop · 2023 · link

TL;DR: We investigate latent space aggregation, merging spaces that differ in both sample and class composition.

Honors & Awards

Professional Activities