← Authors

Sashank Reddi

23 papers · 2015–2023 · 4 conferences · across top CS/AI conferences

Achievements

🐣 Hot Topic Early Bird πŸŒ‰ Interdisciplinary Bridge 🧭 Keyword Pioneer 🌍 Conference Polyglot (4) πŸƒ Academic Marathon (8) 🐝 Cross-Pollinator (12) 🌈 Renaissance Researcher (7) πŸ—ΊοΈ Taxonomy Completionist (42) 🀝 Dynamic Duo (16) πŸ‘‘ Triple Crown πŸ”¬ Deep Specialist (11) πŸ”₯ Unstoppable (6) πŸ’Ž Century Club (23) πŸ—ƒοΈ Keyword Collector (94) ❓ The Questioner (4) ⚑ Prolific Year (7)

Conferences

ICML (9) NIPS (8) AISTATS (3) ICLR (3)

Top co-authors

Sanjiv Kumar (16) Ankit Singh Rawat (8) Srinadh Bhojanapalli (5) Manzil Zaheer (5) Satyen Kale (5) Seungyeon Kim (4) Suvrit Sra (3) Chulhee Yun (3) Aditya K Menon (3) Sai Praneeth Karimireddy (3)

Research topics

Differential Privacy (1)

Keywords

federated learning (4) knowledge distillation (3) adaptive gradient method (3) representation learning (2) information retrieval (2) attention mechanism (2) negative sampling (2) embedding dimension (2) stochastic gradient descent (2) stochastic optimization (2) client drift (2) control variate (2) natural language processing (1) bayesian inference (1) graph laplacian (1) transformer architecture (1) binary classification (1) composite optimization (1) non-convex optimization (1) convergence analysis (1)

Papers

What is the Inductive Bias of Flatness Regularization? A Study of Deep Matrix Factorization Models NIPS 2023
In defense of dual-encoders for neural ranking ICML 2022
Robust Training of Neural Networks Using Scale Invariant Architectures ICML 2022
Private Adaptive Optimization with Side information ICML 2022
Breaking the centralized barrier for cross-device federated learning NIPS 2021
RankDistil: Knowledge Distillation for Ranking AISTATS 2021
Federated Composite Optimization ICML 2021
Disentangling Sampling and Labeling Bias for Learning in Large-output Spaces ICML 2021
A statistical perspective on distillation ICML 2021
Efficient Training of Retrieval Models using Negative Cache NIPS 2021
Learning to Learn by Zeroth-Order Oracle ICLR 2020
O(n) Connections are Expressive Enough: Universal Approximability of Sparse Transformers NIPS 2020
Why are Adaptive Methods Good for Attention Models? NIPS 2020
Are Transformers universal approximators of sequence-to-sequence functions? ICLR 2020
Large Batch Optimization for Deep Learning: Training BERT in 76 minutes ICLR 2020
Low-Rank Bottleneck in Multi-head Attention Models ICML 2020
SCAFFOLD: Stochastic Controlled Averaging for Federated Learning ICML 2020
Escaping Saddle Points with Adaptive Gradient Methods ICML 2019
Multilabel reductions: what is my loss optimising? NIPS 2019
Breaking the Glass Ceiling for Embedding-Based Classifiers for Large Output Spaces NIPS 2019
Adaptive Methods for Nonconvex Optimization NIPS 2018
A Generic Approach for Escaping Saddle points AISTATS 2018
On the High Dimensional Power of a Linear-Time Two Sample Test under Mean-shift Alternatives AISTATS 2015