Atish Agarwala
10 papers · 2021–2025 · 3 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 ↓+2 more ↓ Show less ↑
π Interdisciplinary Bridge Β· Bronze πΊοΈ Taxonomy Completionist (15) π Conference Polyglot Β· Bronze π Keyword Champion (2) β The Questioner
π₯
Unstoppable Β· Bronze
π
Century Club Β· Bronze
Venues
Top co-authors
Research topics
Machine Learning
(10)
Optimization & Theory Β· Machine Learning
(9)
Deep Learning
(5)
Optimization & Theory Β· Deep Learning
(3)
Artificial Intelligence
(2)
Architectures
(2)
Core AI
(2)
Mathematics & Optimization
(1)
Natural Language Processing
(1)
Core Methods
(1)
Learning Types
(1)
Optimization
(1)
Resources & Methods
(1)
Keywords
sharpness-aware minimization
(2)
edge of stability
(2)
neural network
(2)
preconditioning
(2)
hessian eigenvalue
(2)
loss hessian
(2)
neural network optimization
(1)
learning rate
(1)
stochastic gradient descent
(1)
adaptive learning rate
(1)
learning rate decay
(1)
feature learning
(1)
eigenvalue regularization
(1)
high-dimensional analysis
(1)
gradient clipping
(1)
learning dynamics
(1)
training stability
(1)
least square
(1)
symmetric matrix
(1)
multi-task learning
(1)
Papers
Avoiding spurious sharpness minimization broadens applicability of SAM
ICML 2025
Exact risk curves of signSGD in High-Dimensions: quantifying preconditioning and noise-compression effects
ICML 2025
Scaling Collapse Reveals Universal Dynamics in Compute-Optimally Trained Neural Networks
ICML 2025
To Clip or not to Clip: the Dynamics of SGD with Gradient Clipping in High-Dimensions
ICLR 2025
Neglected Hessian component explains mysteries in sharpness regularization
NeurIPS 2024
Stepping on the Edge: Curvature Aware Learning Rate Tuners
NeurIPS 2024
SAM operates far from home: eigenvalue regularization as a dynamical phenomenon
ICML 2023
Second-order regression models exhibit progressive sharpening to the edge of stability
ICML 2023
Deep equilibrium networks are sensitive to initialization statistics
ICML 2022
One Network Fits All? Modular versus Monolithic Task Formulations in Neural Networks
ICLR 2021