Maria Refinetti
7 papers · 2020–2024 · 2 conferences · across top CS/AI conferences
Achievements
Jump to papers ↓+1 more ↓ Show less ↑
🌉 Interdisciplinary Bridge 🧭 Keyword Pioneer 🌍 Conference Polyglot (2) 🐝 Cross-Pollinator (3) 🗺️ Taxonomy Completionist (11)
🔥
Unstoppable
(5)
Conferences
ICML (6)
ICLR (1)
Top co-authors
Keywords
stochastic gradient descent
(2)
double descent
(2)
neural network
(2)
convex optimization
(1)
deep learning theory
(1)
neural network optimization
(1)
high-dimensional statistics
(1)
generalization error
(1)
convex loss
(1)
distribution learning
(1)
generalized linear model
(1)
convolutional network
(1)
gradient alignment
(1)
simplicity bia
(1)
bias variance tradeoff
(1)
principal component
(1)
high-dimensional classification
(1)
generalization theory
(1)
random features regression
(1)
bias variance decomposition
(1)
Papers
Forward Learning with Top-Down Feedback: Empirical and Analytical Characterization
ICLR 2024
Neural networks trained with SGD learn distributions of increasing complexity
ICML 2023
Fluctuations, Bias, Variance & Ensemble of Learners: Exact Asymptotics for Convex Losses in High-Dimension
ICML 2022
The dynamics of representation learning in shallow, non-linear autoencoders
ICML 2022
Classifying high-dimensional Gaussian mixtures: Where kernel methods fail and neural networks succeed
ICML 2021
Align, then memorise: the dynamics of learning with feedback alignment
ICML 2021
Double Trouble in Double Descent: Bias and Variance(s) in the Lazy Regime
ICML 2020