Soham De
12 papers · 2017–2024 · 5 conferences · across top CS/AI conferences
Achievements
Jump to papers ↓+4 more ↓ Show less ↑
π Cross-Pollinator (12) πΊοΈ Taxonomy Completionist (20) π§ Keyword Pioneer π£ Hot Topic Early Bird π Interdisciplinary Bridge
π
Conference Polyglot
(5)
π
Academic Marathon
(7)
π
Century Club
(12)
π
Conference Pioneer
Conferences
ICML (5)
NIPS (3)
ICLR (2)
AISTATS (1)
UAI (1)
Top co-authors
Research topics
Keywords
batch normalization
(3)
stochastic gradient descent
(3)
adversarial robustness
(2)
neural network optimization
(2)
skip connection
(2)
residual network
(2)
non-convex optimization
(1)
formal verification
(1)
model quantization
(1)
recurrent neural network
(1)
adversarial training
(1)
convex optimization
(1)
signal-to-noise ratio
(1)
lagrangian relaxation
(1)
image recognition
(1)
adaptive learning rate
(1)
adversarial attack
(1)
state-space model
(1)
semidefinite programming
(1)
sequence modeling
(1)
Papers
Universality of Linear Recurrences Followed by Non-linear Projections: Finite-Width Guarantees and Benefits of Complex Eigenvalues
ICML 2024
Resurrecting Recurrent Neural Networks for Long Sequences
ICML 2023
High-Performance Large-Scale Image Recognition Without Normalization
ICML 2021
On the Origin of Implicit Regularization in Stochastic Gradient Descent
ICLR 2021
Characterizing signal propagation to close the performance gap in unnormalized ResNets
ICLR 2021
On the Generalization Benefit of Noise in Stochastic Gradient Descent
ICML 2020
Batch Normalization Biases Residual Blocks Towards the Identity Function in Deep Networks
NIPS 2020
The Impact of Neural Network Overparameterization on Gradient Confusion and Stochastic Gradient Descent
ICML 2020
Efficient Neural Network Verification with Exactness Characterization
UAI 2019
Adversarial Robustness through Local Linearization
NIPS 2019
Automated Inference with Adaptive Batches
AISTATS 2017
Training Quantized Nets: A Deeper Understanding
NIPS 2017