Picture for Luca Saglietti

Luca Saglietti

How transformers learn structured data: insights from hierarchical filtering

Add code
Aug 27, 2024
Figure 1 for How transformers learn structured data: insights from hierarchical filtering
Figure 2 for How transformers learn structured data: insights from hierarchical filtering
Figure 3 for How transformers learn structured data: insights from hierarchical filtering
Figure 4 for How transformers learn structured data: insights from hierarchical filtering
Viaarxiv icon

Tilting the Odds at the Lottery: the Interplay of Overparameterisation and Curricula in Neural Networks

Add code
Jun 03, 2024
Figure 1 for Tilting the Odds at the Lottery: the Interplay of Overparameterisation and Curricula in Neural Networks
Figure 2 for Tilting the Odds at the Lottery: the Interplay of Overparameterisation and Curricula in Neural Networks
Figure 3 for Tilting the Odds at the Lottery: the Interplay of Overparameterisation and Curricula in Neural Networks
Figure 4 for Tilting the Odds at the Lottery: the Interplay of Overparameterisation and Curricula in Neural Networks
Viaarxiv icon

The twin peaks of learning neural networks

Add code
Jan 23, 2024
Viaarxiv icon

The star-shaped space of solutions of the spherical negative perceptron

Add code
May 18, 2023
Figure 1 for The star-shaped space of solutions of the spherical negative perceptron
Figure 2 for The star-shaped space of solutions of the spherical negative perceptron
Figure 3 for The star-shaped space of solutions of the spherical negative perceptron
Figure 4 for The star-shaped space of solutions of the spherical negative perceptron
Viaarxiv icon

Optimal transfer protocol by incremental layer defrosting

Add code
Mar 02, 2023
Figure 1 for Optimal transfer protocol by incremental layer defrosting
Figure 2 for Optimal transfer protocol by incremental layer defrosting
Figure 3 for Optimal transfer protocol by incremental layer defrosting
Figure 4 for Optimal transfer protocol by incremental layer defrosting
Viaarxiv icon

Inducing bias is simpler than you think

Add code
May 31, 2022
Figure 1 for Inducing bias is simpler than you think
Figure 2 for Inducing bias is simpler than you think
Figure 3 for Inducing bias is simpler than you think
Figure 4 for Inducing bias is simpler than you think
Viaarxiv icon

An Analytical Theory of Curriculum Learning in Teacher-Student Networks

Add code
Jun 15, 2021
Figure 1 for An Analytical Theory of Curriculum Learning in Teacher-Student Networks
Figure 2 for An Analytical Theory of Curriculum Learning in Teacher-Student Networks
Figure 3 for An Analytical Theory of Curriculum Learning in Teacher-Student Networks
Figure 4 for An Analytical Theory of Curriculum Learning in Teacher-Student Networks
Viaarxiv icon

Probing transfer learning with a model of synthetic correlated datasets

Add code
Jun 09, 2021
Figure 1 for Probing transfer learning with a model of synthetic correlated datasets
Figure 2 for Probing transfer learning with a model of synthetic correlated datasets
Figure 3 for Probing transfer learning with a model of synthetic correlated datasets
Figure 4 for Probing transfer learning with a model of synthetic correlated datasets
Viaarxiv icon

Solvable Model for Inheriting the Regularization through Knowledge Distillation

Add code
Dec 02, 2020
Figure 1 for Solvable Model for Inheriting the Regularization through Knowledge Distillation
Figure 2 for Solvable Model for Inheriting the Regularization through Knowledge Distillation
Figure 3 for Solvable Model for Inheriting the Regularization through Knowledge Distillation
Figure 4 for Solvable Model for Inheriting the Regularization through Knowledge Distillation
Viaarxiv icon

Large deviations for the perceptron model and consequences for active learning

Add code
Dec 09, 2019
Figure 1 for Large deviations for the perceptron model and consequences for active learning
Figure 2 for Large deviations for the perceptron model and consequences for active learning
Figure 3 for Large deviations for the perceptron model and consequences for active learning
Figure 4 for Large deviations for the perceptron model and consequences for active learning
Viaarxiv icon