3 citations · 9 across the 21 of their papers we have counts for
21 papers
A Defense of the Quadratic Model
Alexandru Meterez, Pranav Ajit Nair, Depen Morwani +3
Due to the complexity of neural network loss landscapes, optimization theory is forced to rely on idealized models, and there is generally a tradeoff between how theoretically trac…
Compute Efficiency and Serial Runtime Tradeoffs for Stochastic Momentum Methods
Depen Morwani, Alexandru Meterez, Pranav Nair +1
Stochastic momentum methods such as heavy ball (HB), Nesterov momentum, and variants of Accelerated SGD (ASGD) [Kidambi et al., 2018] are widely used in modern training, but their…
The Recurrent Transformer: Greater Effective Depth and Efficient Decoding
Costin-Andrei Oncescu, Depen Morwani, Samy Jelassi +3
Transformers process tokens in parallel but are temporally shallow: at position , each layer attends to key-value pairs computed based on the previous layer, yielding a depth ca…
GenAI for Systems: Recurring Challenges and Design Principles from Software to Silicon
Arya Tschand, Chenyu Wang, Zishen Wan +21
Generative AI is reshaping how computing systems are designed, optimized, and built, yet research remains fragmented across software, architecture, and chip design communities. Thi…
Anytime Pretraining: Horizon-Free Learning-Rate Schedules with Weight Averaging
Alexandru Meterez, Pranav Ajit Nair, Depen Morwani +2
Large language models are increasingly trained in continual or open-ended settings, where the total training horizon is not known in advance. Despite this, most existing pretrainin…
Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise
Bingbin Liu, Rachit Bansal, Depen Morwani +3
Preconditioned methods are central to deep learning optimization. Predominant approaches include computationally light diagonal preconditioners such as Adam, which rely on gradient…
LOTION: Smoothing the Optimization Landscape for Quantized Training
Mujin Kwun, Depen Morwani, Chloe Huangyuan Su +3
Optimizing neural networks for quantized objectives is fundamentally challenging because the quantizer is piece-wise constant, yielding zero gradients everywhere except at quantiza…
Seesaw: Accelerating Training by Balancing Learning Rate and Batch Size Scheduling
Alexandru Meterez, Depen Morwani, Jingfeng Wu +3
Increasing the batch size during training -- a ''batch ramp'' -- is a promising strategy to accelerate large language model pretraining. While for SGD, doubling the batch size can…
The Potential of Second-Order Optimization for LLMs: A Study with Full Gauss-Newton
Natalie Abreu, Nikhil Vyas, Sham Kakade +1
Recent efforts to accelerate LLM pretraining have focused on computationally-efficient approximations that exploit second-order structure. This raises a key question for large-scal…
A Simplified Analysis of SGD for Linear Regression with Weight Averaging
Alexandru Meterez, Depen Morwani, Costin-Andrei Oncescu +3
Theoretically understanding stochastic gradient descent (SGD) in overparameterized models has led to the development of several optimization algorithms that are widely used in prac…
Connections between Schedule-Free Optimizers, AdEMAMix, and Accelerated SGD Variants
Depen Morwani, Nikhil Vyas, Hanlin Zhang +1
Recent advancements in deep learning optimization have introduced new algorithms, such as Schedule-Free optimizers, AdEMAMix, MARS and Lion which modify traditional momentum mechan…
How Does Critical Batch Size Scale in Pre-training?
Hanlin Zhang, Depen Morwani, Nikhil Vyas +5
Training large-scale models under given resources requires careful design of parallelism strategies. In particular, the efficiency notion of critical batch size (CBS), concerning t…
SOAP: Improving and Stabilizing Shampoo using Adam
Nikhil Vyas, Depen Morwani, Rosie Zhao +5
There is growing evidence of the effectiveness of Shampoo, a higher-order preconditioning method, over Adam in deep learning optimization tasks. However, Shampoo's drawbacks includ…
Deconstructing What Makes a Good Optimizer for Language Models
Rosie Zhao, Depen Morwani, David Brandfonbrener +2
Training language models becomes increasingly expensive with scale, prompting numerous attempts to improve optimization efficiency. Despite these efforts, the Adam optimizer remain…
A New Perspective on Shampoo's Preconditioner
Depen Morwani, Itai Shapira, Nikhil Vyas +3
Shampoo, a second-order optimization algorithm which uses a Kronecker product preconditioner, has recently garnered increasing attention from the machine learning community. The pr…
Feature emergence via margin maximization: case studies in algebraic tasks
Depen Morwani, Benjamin L. Edelman, Costin-Andrei Oncescu +2
Understanding the internal representations learned by neural networks is a cornerstone challenge in the science of machine learning. While there have been significant recent stride…
Beyond Implicit Bias: The Insignificance of SGD Noise in Online Learning
Nikhil Vyas, Depen Morwani, Rosie Zhao +3
The success of SGD in deep learning has been ascribed by prior works to the implicit bias induced by finite batch sizes ("SGD noise"). While prior works focused on offline learning…
Feature-Learning Networks Are Consistent Across Widths At Realistic Scales
Nikhil Vyas, Alexander Atanasov, Blake Bordelon +3
We study the effect of width on the dynamics of feature-learning neural networks across a variety of architectures and datasets. Early in training, wide neural networks trained on…
Simplicity Bias in 1-Hidden Layer Neural Networks
Depen Morwani, Jatin Batra, Prateek Jain +1
Recent works have demonstrated that neural networks exhibit extreme simplicity bias(SB). That is, they learn only the simplest features to solve a task at hand, even in the presenc…
Using noise resilience for ranking generalization of deep neural networks
Depen Morwani, Rahul Vashisht, Harish G. Ramaswamy
Recent papers have shown that sufficiently overparameterized neural networks can perfectly fit even random labels. Thus, it is crucial to understand the underlying reason behind th…
Inductive Bias of Gradient Descent for Weight Normalized Smooth Homogeneous Neural Nets
Depen Morwani, Harish G. Ramaswamy
We analyze the inductive bias of gradient descent for weight normalized smooth homogeneous neural nets, when trained on exponential or cross-entropy loss. We analyse both standard…