4 citations · 13 across the 39 of their papers we have counts for
9 papers · 1 filter
Double Preconditioning (DoPr): Optimization for Test-Time Performance, not Validation Loss
Thomas T. Zhang, Alok Shah, Yifei Zhang +3
Many modern applications of deep learning involve training a neural network via a one-step prediction loss (e.g., regression, cross-entropy), but deploy the network by rollin…
Nearly Instance-Optimal Parameter Recovery from Many Trajectories via Hellinger Localization
Eliot Shekhtman, Yichen Zhou, Ingvar Ziemann +2
Learning from temporally-correlated data is a core facet of modern machine learning. Yet our understanding of sequential learning remains incomplete, particularly in the multi-traj…
Action Chunking and Exploratory Data Collection Yield Exponential Improvements in Behavior Cloning for Continuous Control
Thomas T. Zhang, Daniel Pfrommer, Chaoyi Pan +2
This paper presents a theoretical analysis of two of the most impactful interventions in modern learning from demonstration in robotics and continuous control: the practice of acti…
On The Concurrence of Layer-wise Preconditioning Methods and Provable Feature Learning
Thomas T. Zhang, Behrad Moniri, Ansh Nagwekar +4
Layer-wise preconditioning methods are a family of memory-efficient optimization algorithms that introduce preconditioners per axis of each layer's weight tensors. These methods ha…
Logarithmic Regret for Nonlinear Control
James Wang, Bruce D. Lee, Ingvar Ziemann +1
We address the problem of learning to control an unknown nonlinear dynamical system through sequential interactions. Motivated by high-stakes applications in which mistakes can be…
State space models, emergence, and ergodicity: How many parameters are needed for stable predictions?
Ingvar Ziemann, Nikolai Matni, George J. Pappas
How many parameters are required for a model to execute a given task? It has been argued that large language models, pre-trained via self-supervised learning, exhibit emergent capa…
Regret Analysis of Multi-task Representation Learning for Linear-Quadratic Adaptive Control
Bruce D. Lee, Leonardo F. Toso, Thomas T. Zhang +2
Representation learning is a powerful tool that enables learning over large multitudes of agents or domains by enforcing that all agents operate on a shared set of learned features…
Single Trajectory Conformal Prediction
Brian Lee, Nikolai Matni
We study the performance of risk-controlling prediction sets (RCPS), an empirical risk minimization-based formulation of conformal prediction, with a single trajectory of temporall…
Sharp Rates in Dependent Learning Theory: Avoiding Sample Size Deflation for the Square Loss
Ingvar Ziemann, Stephen Tu, George J. Pappas +1
In this work, we study statistical learning with dependent (-mixing) data and square loss in a hypothesis class where is the norm $\|f\|_{Ψ_p}…