SR
Sashank J. Reddi
cs.LGstat.MLmath.OCcs.AIcs.CLcs.DCmath.STstat.THcs.CRcs.IT
On Valency
published · living versionsW_nkj8q39j·v1 · currentpublished
Preprints & journals
54 papers in the corpus · 2012–2025Bipartite Ranking From Multiple Labels: On Loss Versus Label Aggregation2504.11284v2 · Michal Lukasik, Lin Chen, Harikrishna Narasimhan et al.2025 · 0 citationsarXiv
Structured Preconditioners in Adaptive Optimization: A Unified Analysis2503.10537v2 · Shuo Xie, Tianhao Wang, Sashank Reddi et al.2025 · 0 citationsarXiv
Reasoning with Latent Thoughts: On the Power of Looped Transformers2502.17416v1 · Nikunj Saunshi, Nishanth Dikkala, Zhiyuan Li et al.2025 · 2 citationsarXiv
On the Role of Depth and Looping for In-Context Learning with Task Diversity2410.21698v1 · Khashayar Gatmiry, Nikunj Saunshi, Sashank J. Reddi et al.2024 · 0 citationsarXiv
A Little Help Goes a Long Way: Efficient LLM Training by Leveraging Small LMs2410.18779v1 · Ankit Singh Rawat, Veeranjaneyulu Sadhanala, Afshin Rostamizadeh et al.2024 · 1 citationarXiv
Simplicity Bias via Global Convergence of Sharpness Minimization2410.16401v1 · Khashayar Gatmiry, Zhiyuan Li, Sashank J. Reddi et al.2024 · 0 citationsarXiv
Can Looped Transformers Learn to Implement Multi-step Gradient Descent for In-context Learning?2410.08292v1 · Khashayar Gatmiry, Nikunj Saunshi, Sashank J. Reddi et al.2024 · 0 citationsarXiv
On the Inductive Bias of Stacking Towards Improving Reasoning2409.19044v1 · Nikunj Saunshi, Stefani Karp, Shankar Krishnan et al.2024 · 1 citationarXiv
Efficient Document Ranking with Learnable Late Interactions2406.17968v1 · Ziwei Ji, Himanshu Jain, Andreas Veit et al.2024 · 3 citationsarXiv
Landscape-Aware Growing: The Power of a Little LAG2406.02469v1 · Stefani Karp, Nikunj Saunshi, Sobhan Miryoosefi et al.2024 · 0 citationsarXiv
Efficient Stagewise Pretraining via Progressive Subnetworks2402.05913v2 · Abhishek Panigrahi, Nikunj Saunshi, Kaifeng Lyu et al.2024 · 0 citationsarXiv
The Inductive Bias of Flatness Regularization for Deep Matrix Factorization2306.13239v1 · Khashayar Gatmiry, Zhiyuan Li, Ching-Yao Chuang et al.2023 · 1 citationarXiv
Depth Dependence of P Learning Rates in ReLU MLPs2305.07810v1 · Samy Jelassi, Boris Hanin, Ziwei Ji et al.2023 · 0 citationsarXiv
Differentially Private Adaptive Optimization with Delayed Preconditioners2212.00309v2 · Tian Li, Manzil Zaheer, Ken Ziyu Liu et al.2022 · 3 citationsarXiv
On the Algorithmic Stability and Generalization of Adaptive Optimization Methods2211.03970v1 · Han Nguyen, Hai Pham, Sashank J. Reddi et al.2022 · 0 citationsarXiv
The Lazy Neuron Phenomenon: On Emergence of Activation Sparsity in Transformers2210.06313v2 · Zonglin Li, Chong You, Srinadh Bhojanapalli et al.2022 · 18 citationsarXiv
Private Adaptive Optimization with Side Information2202.05963v2 · Tian Li, Manzil Zaheer, Sashank J. Reddi et al.2022 · 8 citationsarXiv
Robust Training of Neural Networks Using Scale Invariant Architectures2202.00980v2 · Zhiyuan Li, Srinadh Bhojanapalli, Manzil Zaheer et al.2022 · 3 citationsarXiv
FedLite: A Scalable Approach for Federated Learning on Resource-constrained Clients2201.11865v2 · Jianyu Wang, Hang Qi, Ankit Singh Rawat et al.2022 · 14 citationsarXiv
A Field Guide to Federated Optimization2107.06917v1 · Jianyu Wang, Zachary Charles, Zheng Xu et al.2021 · 181 citationsarXiv
Disentangling Sampling and Labeling Bias for Learning in Large-Output Spaces2105.05736v1 · Ankit Singh Rawat, Aditya Krishna Menon, Wittawat Jitkrittum et al.2021 · 5 citationsarXiv
Distilling Double Descent2102.06849v1 · Andrew Cotter, Aditya Krishna Menon, Harikrishna Narasimhan et al.2021 · 0 citationsarXiv
Federated Composite Optimization2011.08474v3 · Honglin Yuan, Manzil Zaheer, Sashank Reddi2020 · 29 citationsarXiv
Mime: Mimicking Centralized Stochastic Algorithms in Federated Learning2008.03606v2 · Sai Praneeth Karimireddy, Martin Jaggi, Satyen Kale et al.2020 · 96 citationsarXiv
Connections are Expressive Enough: Universal Approximability of Sparse Transformers2006.04862v2 · Chulhee Yun, Yin-Wen Chang, Srinadh Bhojanapalli et al.2020 · 33 citationsarXiv
Why distillation helps: a statistical perspective2005.10419v1 · Aditya Krishna Menon, Ankit Singh Rawat, Sashank J. Reddi et al.2020 · 12 citationsarXiv
Doubly-stochastic mining for heterogeneous retrieval2004.10915v1 · Ankit Singh Rawat, Aditya Krishna Menon, Andreas Veit et al.2020 · 3 citationsarXiv
Adaptive Federated Optimization2003.00295v5 · Sashank Reddi, Zachary Charles, Manzil Zaheer et al.2020 · 840 citationsarXiv
Adaptive Sampling Distributed Stochastic Variance Reduced Gradient for Heterogeneous Distributed Datasets2002.08528v3 · Ilqar Ramazanli, Han Nguyen, Hai Pham et al.2020 · 6 citationsarXiv
Low-Rank Bottleneck in Multi-head Attention Models2002.07028v1 · Srinadh Bhojanapalli, Chulhee Yun, Ankit Singh Rawat et al.2020 · 45 citationsarXiv
Are Transformers universal approximators of sequence-to-sequence functions?1912.10077v2 · Chulhee Yun, Srinadh Bhojanapalli, Ankit Singh Rawat et al.2019 · 120 citationsarXiv
Why are Adaptive Methods Good for Attention Models?1912.03194v2 · Jingzhao Zhang, Sai Praneeth Karimireddy, Andreas Veit et al.2019 · 70 citationsarXiv
Learning to Learn by Zeroth-Order Oracle1910.09464v2 · Yangjun Ruan, Yuanhao Xiong, Sashank Reddi et al.2019 · 8 citationsarXiv
SCAFFOLD: Stochastic Controlled Averaging for Federated Learning1910.06378v4 · Sai Praneeth Karimireddy, Satyen Kale, Mehryar Mohri et al.2019 · 1,733 citationsarXiv
AdaCliP: Adaptive Clipping for Private SGD1908.07643v2 · Venkatadheeraj Pichapati, Ananda Theertha Suresh, Felix X. Yu et al.2019 · 63 citationsarXiv
On the Convergence of Adam and Beyond1904.09237v1 · Sashank J. Reddi, Satyen Kale, Sanjiv Kumar2019 · 1,771 citationsarXivon Valency
Large Batch Optimization for Deep Learning: Training BERT in 76 minutes1904.00962v5 · Yang You, Jing Li, Sashank Reddi et al.2019 · 473 citationsarXiv
Escaping Saddle Points with Adaptive Gradient Methods1901.09149v2 · Matthew Staib, Sashank J. Reddi, Satyen Kale et al.2019 · 49 citationsarXiv
Stochastic Negative Mining for Learning with Large Output Spaces1810.07076v1 · Sashank J. Reddi, Satyen Kale, Felix Yu et al.2018 · 30 citationsarXiv
A Generic Approach for Escaping Saddle points1709.01434v1 · Sashank J Reddi, Manzil Zaheer, Suvrit Sra et al.2017 · 54 citationsarXiv
Variance Reduction in Stochastic Gradient Langevin Dynamics.28713210 · Dubey, Avinava, Reddi, Sashank J, Póczos, Barnabás et al.2019 · 65 citationsAdvances in neural information processing systems. 2016;29:1154-1162
AIDE: Fast and Communication Efficient Distributed Optimization1608.06879v1 · Sashank J. Reddi, Jakub Konecn'y, Peter Richt'arik et al.2016 · 98 citationsarXiv
Stochastic Frank-Wolfe Methods for Nonconvex Optimization1607.08254v2 · Sashank J. Reddi, Suvrit Sra, Barnabas Poczos et al.2016 · 149 citationsarXiv
Fast Stochastic Methods for Nonsmooth Nonconvex Optimization1605.06900v1 · Sashank J. Reddi, Suvrit Sra, Barnabas Poczos et al.2016 · 34 citationsarXiv
Riemannian SVRG: Fast Stochastic Optimization on Riemannian Manifolds1605.07147v2 · Hongyi Zhang, Sashank J. Reddi, Suvrit Sra2016 · 131 citationsAdvances in Neural Information Processing Systems 29 (NIPS 2016)
Fast Incremental Method for Nonconvex Optimization1603.06159v1 · Sashank J. Reddi, Suvrit Sra, Barnabas Poczos et al.2016 · 27 citationsarXiv
Stochastic Variance Reduction for Nonconvex Optimization1603.06160v2 · Sashank J. Reddi, Ahmed Hefny, Suvrit Sra et al.2016 · 385 citationsarXiv
Adaptivity and Computation-Statistics Tradeoffs for Kernel and Distance based High Dimensional Two Sample Testing1508.00655v1 · Aaditya Ramdas, Sashank J. Reddi, Barnabas Poczos et al.2015 · 17 citationsarXiv
On Variance Reduction in Stochastic Gradient Descent and its Asynchronous Variants1506.06840v2 · Sashank J. Reddi, Ahmed Hefny, Suvrit Sra et al.2015 · 133 citationsarXiv
On the High-dimensional Power of Linear-time Kernel Two-Sample Testing under Mean-difference Alternatives1411.6314v1 · Aaditya Ramdas, Sashank J. Reddi, Barnabas Poczos et al.2014 · 5 citationsarXiv
Career total: 81 works. 54 are in this corpus.Showing the 50 most recent.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.