Antonio Orvieto

12

Papers

3

Total Citations

Papers (12)

Geometric Inductive Biases of Deep Networks: The Role of Data and Architecture

Universality of Linear Recurrences Followed by Non-linear Projections: Finite-Width Guarantees and Benefits of Complex Eigenvalues

Achieving a Better Stability-Plasticity Trade-Off via Auxiliary Networks in Continual Learning

Recurrent Distance Filtering for Graph Representation Learning

Enhancing Optimizer Stability: Momentum Adaptation of The NGN Step-size

On the Theoretical Properties of Noise Correlation in Stochastic Optimization

Dynamics of SGD with Stochastic Polyak Stepsizes: Truly Adaptive Variants and Convergence to Exact Solution

Signal Propagation in Transformers: Theoretical Perspectives and the Role of Rank Collapse

Shadowing Properties of Optimization Algorithms

Continuous-time Models for Stochastic Optimization Algorithms

Rethinking the Variational Interpretation of Accelerated Optimization Methods

On the Second-order Convergence Properties of Random Search Methods