Peter Súkeník
5 papers · 2022–2025 · 2 venues · across top CS/AI venues
Authors are matched by name, so researchers who share a name ("Yang Liu", "Wei Wang") appear as a single profile and their paper counts, venues and achievements are combined. How identity is handled.
Achievements
Jump to papers ↓
🌍
Conference Polyglot · Bronze
❓
The Questioner
Venues
Top co-authors
Research topics
Keywords
gradient descent
(3)
neural collapse
(3)
deep neural network
(2)
binary classification
(1)
knowledge distillation
(1)
multi-class classification
(1)
singular value decomposition
(1)
convolutional neural network
(1)
weight decay
(1)
wide neural network
(1)
fully connected layer
(1)
global optimum
(1)
unconstrained features model
(1)
neural network
(1)
low-data regime
(1)
deep neural collapse
(1)
gradient outer product
(1)
low-rank bia
(1)
balancedness condition
(1)
active learning
(1)
Papers
Wide Neural Networks Trained with Weight Decay Provably Exhibit Neural Collapse
ICLR 2025
Average gradient outer product as a mechanism for deep neural collapse
NeurIPS 2024
Neural collapse vs. low-rank bias: Is deep neural collapse really optimal?
NeurIPS 2024
Deep Neural Collapse Is Provably Optimal for the Deep Unconstrained Features Model
NeurIPS 2023
The Unreasonable Effectiveness of Fully-Connected Layers for Low-Data Regimes
NeurIPS 2022