Liu Ziyin
17 papers · 2020–2025 · 3 conferences · across top CS/AI conferences
Achievements
Jump to papers ↓+6 more ↓ Show less ↑
🐝 Cross-Pollinator (8) 🏃 Academic Marathon (5) 🌍 Conference Polyglot (3) 🧭 Keyword Pioneer 🌉 Interdisciplinary Bridge
🌉
Interdisciplinary Bridge
👑
Triple Crown
⚡
Prolific Year
(5)
💎
Century Club
(17)
❓
The Questioner
🔥
Unstoppable
(6)
Conferences
ICML (7)
ICLR (6)
NIPS (4)
Top co-authors
Keywords
stochastic gradient descent
(4)
escape rate
(2)
neural network optimization
(2)
representation learning
(2)
learning rate
(1)
embedding learning
(1)
feature selection
(1)
weight decay
(1)
optimization theory
(1)
stationary distribution
(1)
learning dynamics
(1)
latent variable model
(1)
inductive bia
(1)
hessian eigenvalue
(1)
implicit bia
(1)
variational autoencoder
(1)
self-supervised learning
(1)
stochastic differential equation
(1)
contrastive learning
(1)
neural network compression
(1)
Papers
Understanding the Emergence of Multimodal Representation Alignment
ICML 2025
Three Mechanisms of Feature Learning in a Linear Network
ICLR 2025
Formation of Representations in Neural Networks
ICLR 2025
Remove Symmetries to Control Model Expressivity and Improve Optimization
ICLR 2025
Compositional Generalization via Forced Rendering of Disentangled Latents
ICML 2025
Symmetry Induces Structure and Constraint of Learning
ICML 2024
Parameter Symmetry and Noise Equilibrium of Stochastic Gradient Descent
NIPS 2024
On the Stepwise Nature of Self-Supervised Learning
ICML 2023
What shapes the loss landscape of self supervised learning?
ICLR 2023
spred: Solving L1 Penalty with SGD
ICML 2023
Strength of Minibatch Noise in SGD
ICLR 2022
Power-Law Escape Rate of SGD
ICML 2022
Posterior Collapse of a Linear Latent Variable Model
NIPS 2022
SGD Can Converge to Local Maxima
ICLR 2022
Exact Solutions of a Deep Linear Network
NIPS 2022
Noise and Fluctuation of Finite Learning Rate Stochastic Gradient Descent
ICML 2021
Neural Networks Fail to Learn Periodic Functions and How to Fix It
NIPS 2020