AI safety (1)
cognitive control (1)
compositionality (1)
continual learning (4)
curriculum learning (5)
data imbalance (3)
diffusion models (1)
epidemic mitigation (1)
fairness (3)
fatigue (1)
landscape (7)
large language models (1)
lottery ticket hypothesis (1)
machine learning (1)
model collapse (1)
momentum (1)
neural networks (1)
optimal control (1)
optimisation (8)
reinforcement learning (1)
review (2)
spurious correlations (1)
statistical physics (1)
transfer learning (2)
AI safety (1)
Position: the Stochastic Parrot in the Coal Mine. Model Collapse is a Threat to Low-Resource Communities
Devon Jarvis, Richard Klein, Benjamin Rosman, Steven James, Stefano Sarao Mannelli
ICML 2026 (Spotlight)
cognitive control (1)
A meta-learning framework for rationalizing cognitive fatigue in neural systems
Yujun Li, Rodrigo Carrasco-Davis, Younes Strittmatter, Stefano Sarao Mannelli, Sebastian Musslick
CogSci 2024 (Oral)
compositionality (1)
Why Do Animals Need Shaping? A Theory of Task Composition and Curriculum Learning
Jin Hwa Lee, Stefano Sarao Mannelli, Andrew Saxe
ICML 2024
continual learning (4)
Maslow's Hammer for Catastrophic Forgetting: Node Re-Use vs Node Activation
Sebastian Lee, Stefano Sarao Mannelli, Claudia Clopath, Sebastian Goldt, Andrew Saxe
ICML 2022
A meta-learning framework for rationalizing cognitive fatigue in neural systems
Yujun Li, Rodrigo Carrasco-Davis, Younes Strittmatter, Stefano Sarao Mannelli, Sebastian Musslick
CogSci 2024 (Oral)
Optimal Protocols for Continual Learning via Statistical Physics and Control Theory
Francesco Mori, Stefano Sarao Mannelli, Francesca Mignacco
ICLR 2025; J. Stat. Mech. 2025, 084004
A Theory of Initialisation's Impact on Specialisation
Devon Jarvis, Sebastian Lee, Clémentine Carla Juliette Dominé, Andrew M Saxe, Stefano Sarao Mannelli
ICLR 2025; J. Stat. Mech. 2025, 114001
curriculum learning (5)
An Analytical Theory of Curriculum Learning in Teacher-Student Networks
Luca Saglietti*, Stefano Sarao Mannelli*, Andrew Saxe
NeurIPS 2022
RL Perceptron: Generalization Dynamics of Policy Learning in High Dimensions
Nishil Patel, Sebastian Lee, Stefano Sarao Mannelli, Sebastian Goldt, Andrew Saxe
Phys. Rev. X 15, 021051 (2025)
Why Do Animals Need Shaping? A Theory of Task Composition and Curriculum Learning
Jin Hwa Lee, Stefano Sarao Mannelli, Andrew Saxe
ICML 2024
Tilting the Odds at the Lottery: the Interplay of Overparameterisation and Curricula in Neural Networks
Stefano Sarao Mannelli, Yaraslau Ivashinka, Andrew Saxe, Luca Saglietti
ICML 2024
Curriculum learning in humans and neural networks
Younes Strittmatter*, Stefano Sarao Mannelli*, Miguel Ruiz-Garcia, Sebastian Musslick, Markus Wolfgang Hermann Spitzer
Proceedings of the Annual Meeting of the Cognitive Science Society 47 (CogSci 2025)
data imbalance (3)
Bias in Motion: Theoretical Insights into the Dynamics of Bias in SGD Training
Anchit Jain, Rozhin Nobahari, Aristide Baratin, Stefano Sarao Mannelli
NeurIPS 2024
Bias-inducing geometries: an exactly solvable data model with fairness implications
Stefano Sarao Mannelli, Federica Gerace, Negar Rostamzadeh, Luca Saglietti
Phys. Rev. E 112, 025304 (2025)
The Interplay of Data Structure and Imbalance in the Learning Dynamics of Diffusion Models
Flavio Nicoletti, Chenxiao Ma, Enrico Ventura, Luca Saglietti, Stefano Sarao Mannelli
arXiv preprint
diffusion models (1)
The Interplay of Data Structure and Imbalance in the Learning Dynamics of Diffusion Models
Flavio Nicoletti, Chenxiao Ma, Enrico Ventura, Luca Saglietti, Stefano Sarao Mannelli
arXiv preprint
epidemic mitigation (1)
Epidemic mitigation by statistical inference from contact tracing data
Antoine Baker, Indaco Biazzo, Alfredo Braunstein, Giovanni Catania, Luca Dall’Asta, Alessandro Ingrosso, Florent Krzakala, Fabio Mazza, Marc Mezard, Anna Paola Muntoni, Maria Refinetti, Stefano Sarao Mannelli, Lenka Zdeborova
Proceedings of the National Academy of Sciences
fairness (3)
Bias in Motion: Theoretical Insights into the Dynamics of Bias in SGD Training
Anchit Jain, Rozhin Nobahari, Aristide Baratin, Stefano Sarao Mannelli
NeurIPS 2024
Bias-inducing geometries: an exactly solvable data model with fairness implications
Stefano Sarao Mannelli, Federica Gerace, Negar Rostamzadeh, Luca Saglietti
Phys. Rev. E 112, 025304 (2025)
The Interplay of Data Structure and Imbalance in the Learning Dynamics of Diffusion Models
Flavio Nicoletti, Chenxiao Ma, Enrico Ventura, Luca Saglietti, Stefano Sarao Mannelli
arXiv preprint
fatigue (1)
A meta-learning framework for rationalizing cognitive fatigue in neural systems
Yujun Li, Rodrigo Carrasco-Davis, Younes Strittmatter, Stefano Sarao Mannelli, Sebastian Musslick
CogSci 2024 (Oral)
landscape (7)
Passed & Spurious: Descent Algorithms and Local Minima in Spiked Matrix-Tensor Models
Stefano Sarao Mannelli, Florent Krzakala, Pierfrancesco Urbani, Lenka Zdeborova
ICML 2019
Who is Afraid of Big Bad Minima? Analysis of gradient-flow in spiked matrix-tensor models
Stefano Sarao Mannelli, Giulio Biroli, Chiara Cammarota, Florent Krzakala, Lenka Zdeborova
NeurIPS 2019 (Spotlight)
Marvels and pitfalls of the Langevin algorithm in noisy high-dimensional inference
Stefano Sarao Mannelli, Giulio Biroli, Chiara Cammarota, Florent Krzakala, Pierfrancesco Urbani, Lenka Zdeborova
Physical Review X
Thresholds of descending algorithms in inference problems
Stefano Sarao Mannelli, Lenka Zdeborova
Journal of Statistical Mechanics: Theory and Experiment
Complex Dynamics in Simple Neural Networks: Understanding Gradient Flow in Phase Retrieval
Stefano Sarao Mannelli, Giulio Biroli, Chiara Cammarota, Florent Krzakala, Pierfrancesco Urbani, Lenka Zdeborova
NeurIPS 2020
Optimization and Generalization of Shallow Neural Networks with Quadratic Activation Functions
Stefano Sarao Mannelli, Eric Vanden-Eijnden, Lenka Zdeborova
NeurIPS 2020
large language models (1)
Position: the Stochastic Parrot in the Coal Mine. Model Collapse is a Threat to Low-Resource Communities
Devon Jarvis, Richard Klein, Benjamin Rosman, Steven James, Stefano Sarao Mannelli
ICML 2026 (Spotlight)
lottery ticket hypothesis (1)
Tilting the Odds at the Lottery: the Interplay of Overparameterisation and Curricula in Neural Networks
Stefano Sarao Mannelli, Yaraslau Ivashinka, Andrew Saxe, Luca Saglietti
ICML 2024
machine learning (1)
Thinking of Neural Networks Like a Physicist: The Statistical Physics of Machine Learning
Kai Jappe Sandbrink, Stefano Sarao Mannelli, Florent Krzakala
Proceedings of the Analytical Connectionism Schools 2023--2024, PMLR 320:15-41, 2026
model collapse (1)
Position: the Stochastic Parrot in the Coal Mine. Model Collapse is a Threat to Low-Resource Communities
Devon Jarvis, Richard Klein, Benjamin Rosman, Steven James, Stefano Sarao Mannelli
ICML 2026 (Spotlight)
momentum (1)
Analytical Study of Momentum-Based Acceleration Methods in Paradigmatic High-Dimensional Non-Convex Problems
Stefano Sarao Mannelli, Pierfrancesco Urbani
NeurIPS 2021
neural networks (1)
optimal control (1)
Optimal Protocols for Continual Learning via Statistical Physics and Control Theory
Francesco Mori, Stefano Sarao Mannelli, Francesca Mignacco
ICLR 2025; J. Stat. Mech. 2025, 084004
optimisation (8)
Passed & Spurious: Descent Algorithms and Local Minima in Spiked Matrix-Tensor Models
Stefano Sarao Mannelli, Florent Krzakala, Pierfrancesco Urbani, Lenka Zdeborova
ICML 2019
Who is Afraid of Big Bad Minima? Analysis of gradient-flow in spiked matrix-tensor models
Stefano Sarao Mannelli, Giulio Biroli, Chiara Cammarota, Florent Krzakala, Lenka Zdeborova
NeurIPS 2019 (Spotlight)
Marvels and pitfalls of the Langevin algorithm in noisy high-dimensional inference
Stefano Sarao Mannelli, Giulio Biroli, Chiara Cammarota, Florent Krzakala, Pierfrancesco Urbani, Lenka Zdeborova
Physical Review X
Thresholds of descending algorithms in inference problems
Stefano Sarao Mannelli, Lenka Zdeborova
Journal of Statistical Mechanics: Theory and Experiment
Complex Dynamics in Simple Neural Networks: Understanding Gradient Flow in Phase Retrieval
Stefano Sarao Mannelli, Giulio Biroli, Chiara Cammarota, Florent Krzakala, Pierfrancesco Urbani, Lenka Zdeborova
NeurIPS 2020
Optimization and Generalization of Shallow Neural Networks with Quadratic Activation Functions
Stefano Sarao Mannelli, Eric Vanden-Eijnden, Lenka Zdeborova
NeurIPS 2020
Analytical Study of Momentum-Based Acceleration Methods in Paradigmatic High-Dimensional Non-Convex Problems
Stefano Sarao Mannelli, Pierfrancesco Urbani
NeurIPS 2021
reinforcement learning (1)
RL Perceptron: Generalization Dynamics of Policy Learning in High Dimensions
Nishil Patel, Sebastian Lee, Stefano Sarao Mannelli, Sebastian Goldt, Andrew Saxe
Phys. Rev. X 15, 021051 (2025)
review (2)
Thresholds of descending algorithms in inference problems
Stefano Sarao Mannelli, Lenka Zdeborova
Journal of Statistical Mechanics: Theory and Experiment
Thinking of Neural Networks Like a Physicist: The Statistical Physics of Machine Learning
Kai Jappe Sandbrink, Stefano Sarao Mannelli, Florent Krzakala
Proceedings of the Analytical Connectionism Schools 2023--2024, PMLR 320:15-41, 2026
spurious correlations (1)
Bias in Motion: Theoretical Insights into the Dynamics of Bias in SGD Training
Anchit Jain, Rozhin Nobahari, Aristide Baratin, Stefano Sarao Mannelli
NeurIPS 2024
statistical physics (1)
Thinking of Neural Networks Like a Physicist: The Statistical Physics of Machine Learning
Kai Jappe Sandbrink, Stefano Sarao Mannelli, Florent Krzakala
Proceedings of the Analytical Connectionism Schools 2023--2024, PMLR 320:15-41, 2026
transfer learning (2)
Probing transfer learning with a model of synthetic correlated datasets
Federica Gerace, Luca Saglietti, Stefano Sarao Mannelli, Andrew Saxe, Lenka Zdeborová
Machine Learning: Science and Technology
How to choose the right transfer learning protocol? A qualitative analysis in a controlled set-up
Federica Gerace, Diego Doimo, Stefano Sarao Mannelli, Luca Saglietti, Alessandro Laio
TMLR