Felix Dangel
12 papers · 2020–2025 · 4 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 ↓+4 more ↓ Show less ↑
🌉 Interdisciplinary Bridge · Bronze 🗺️ Taxonomy Completionist (18) 🌍 Conference Polyglot · Bronze 🏃 Academic Marathon · Sprint 🏆 Keyword Champion (3)
🗃️
Keyword Collector
(59)
❓
The Questioner
(3)
⚡
Prolific Year · Productive
💎
Century Club · Bronze
Venues
Top co-authors
Research topics
Machine Learning
(10)
Optimization & Theory · Machine Learning
(10)
Deep Learning
(9)
Architectures
(5)
Mathematics & Optimization
(3)
Application Areas
(3)
Optimization & Theory · Deep Learning
(3)
Mathematics
(2)
Techniques
(2)
Artificial Intelligence
(1)
Reinforcement Learning
(1)
Applications
(1)
Core AI
(1)
Core Methods
(1)
Methods
(1)
Keywords
neural network optimization
(4)
curvature approximation
(3)
automatic differentiation
(2)
natural gradient
(1)
generalization
(1)
natural gradient descent
(1)
model interpretability
(1)
fisher information matrix
(1)
riemannian geometry
(1)
second-order optimization
(1)
second-order information
(1)
convolutional neural network
(1)
adam optimizer
(1)
physics-informed neural network
(1)
gauss-newton method
(1)
hessian approximation
(1)
adaptive gradient method
(1)
generalization gap
(1)
computational cost
(1)
attention mechanism
(1)
Papers
Fishers for Free? Approximating the Fisher Information Matrix by Recycling the Squared Gradient Accumulator
ICML 2025
Hide & Seek: Transformer Symmetries Obscure Sharpness & Riemannian Geometry Finds It
ICML 2025
What Does It Mean to Be a Transformer? Insights from a Theoretical Hessian Analysis
ICLR 2025
Can We Remove the Square-Root in Adaptive Gradient Methods? A Second-Order Perspective
ICML 2024
Convolutions and More as Einsum: A Tensor Network Perspective with Advances for Second-Order Methods
NeurIPS 2024
Kronecker-Factored Approximate Curvature for Physics-Informed Neural Networks
NeurIPS 2024
Revisiting Scalable Hessian Diagonal Approximations for Applications in Reinforcement Learning
ICML 2024
Structured Inverse-Free Natural Gradient Descent: Memory-Efficient & Numerically-Stable KFAC
ICML 2024
The Geometry of Neural Nets' Parameter Spaces Under Reparametrization
NeurIPS 2023
Cockpit: A Practical Debugging Tool for the Training of Deep Neural Networks
NeurIPS 2021
BackPACK: Packing more into Backprop
ICLR 2020
Modular Block-diagonal Curvature Approximations for Feedforward Architectures
AISTATS 2020