16 citations · 73 across the 29 of their papers we have counts for
20 papers · 1 filter
Conditional Flow Matching for Transport Between Markov Processes
Syamantak Kumar, Dheeraj Nagaraj, Saptarshi Roy +1
Motivated by sequence-to-sequence transport in the context time-series domain adaptation, we study the problem of transportation between trajectories of Markov processes. Given a l…
Fine-Tuning Diffusion Models via Intermediate Distribution Shaping
Gautham Govind Anil, Shaan Ul Haque, Nithish Kannen +3
Diffusion models are widely used for generative tasks across domains. Given a pre-trained diffusion model, it is often desirable to fine-tune it further either to correct for error…
Beyond Propagation of Chaos: A Stochastic Algorithm for Mean Field Optimization
Chandan Tankala, Dheeraj M. Nagaraj, Anant Raj
Gradient flow in the 2-Wasserstein space is widely used to optimize functionals over probability distributions and is typically implemented using an interacting particle system wit…
Interleaved Gibbs Diffusion: Generating Discrete-Continuous Data with Implicit Constraints
Gautham Govind Anil, Sachin Yadav, Dheeraj Nagaraj +2
We introduce Interleaved Gibbs Diffusion (IGD), a novel generative modeling framework for discrete-continuous data, focusing on problems with important, implicit and unspecified co…
Dimension-free Score Matching and Time Bootstrapping for Diffusion Models
Syamantak Kumar, Dheeraj Nagaraj, Purnamrita Sarkar
Diffusion models generate samples by estimating the score function of the target distribution at various noise levels. The model is trained using samples drawn from the target dist…
The Bandit Whisperer: Communication Learning for Restless Bandits
Yunfan Zhao, Tonghan Wang, Dheeraj Nagaraj +2
Applying Reinforcement Learning (RL) to Restless Multi-Arm Bandits (RMABs) offers a promising avenue for addressing allocation problems with resource constraints and temporal dynam…
The Poisson Midpoint Method for Langevin Dynamics: Provably Efficient Discretization for Diffusion Models
Saravanan Kandasamy, Dheeraj Nagaraj
Langevin Dynamics is a Stochastic Differential Equation (SDE) central to sampling and generative modeling and is implemented via time discretization. Langevin Monte Carlo (LMC), ba…
Glauber Generative Model: Discrete Diffusion Models via Binary Classification
Harshit Varma, Dheeraj Nagaraj, Karthikeyan Shanmugam
We introduce the Glauber Generative Model (GGM), a new class of discrete diffusion models, to obtain new samples from a distribution given samples from a discrete space. GGM deploy…
Towards a Pretrained Model for Restless Bandits via Multi-arm Generalization
Yunfan Zhao, Nikhil Behari, Edward Hughes +5
Restless multi-arm bandits (RMABs), a class of resource allocation problems with broad application in areas such as healthcare, online advertising, and anti-poaching, have recently…
Stochastic Re-weighted Gradient Descent via Distributionally Robust Optimization
Ramnath Kumar, Kushal Majmundar, Dheeraj Nagaraj +1
We present Re-weighted Gradient Descent (RGD), a novel optimization technique that improves the performance of deep neural networks through dynamic sample re-weighting. Leveraging…
Finite time analysis of temporal difference learning with linear function approximation: Tail averaging and regularisation
Gandharv Patil, Prashanth L. A., Dheeraj Nagaraj +1
We study the finite-time behaviour of the popular temporal difference (TD) learning algorithm when combined with tail-averaging. We derive finite time bounds on the parameter error…
Multi-User Reinforcement Learning with Low Rank Rewards
Naman Agarwal, Prateek Jain, Suhas Kowshik +2
In this work, we consider the problem of collaborative multi-user reinforcement learning. In this setting there are multiple users with the same state-action space and transition p…
Introspective Experience Replay: Look Back When Surprised
Ramnath Kumar, Dheeraj Nagaraj
In reinforcement learning (RL), experience replay-based sampling techniques play a crucial role in promoting convergence by eliminating spurious correlations. However, widely used…
Online Target Q-learning with Reverse Experience Replay: Efficiently finding the Optimal Policy for Linear MDPs
Naman Agarwal, Syomantak Chaudhuri, Prateek Jain +2
Q-learning is a popular Reinforcement Learning (RL) algorithm which is widely used in practice with function approximation (Mnih et al., 2015). In contrast, existing theoretical re…
The staircase property: How hierarchical structure can guide deep learning
Emmanuel Abbe, Enric Boix-Adsera, Matthew Brennan +2
This paper identifies a structural property of data distributions that enables deep neural networks to learn hierarchically. We define the "staircase" property for functions over t…
Near-optimal Offline and Streaming Algorithms for Learning Non-Linear Dynamical Systems
Prateek Jain, Suhas S Kowshik, Dheeraj Nagaraj +1
We consider the setting of vector valued non-linear dynamical systems , where is unbiased noise and is a known link…
Streaming Linear System Identification with Reverse Experience Replay
Prateek Jain, Suhas S Kowshik, Dheeraj Nagaraj +1
We consider the problem of estimating a linear time-invariant (LTI) dynamical system from a single trajectory via streaming algorithms, which is encountered in several applications…
A law of robustness for two-layers neural networks
Sébastien Bubeck, Yuanzhi Li, Dheeraj Nagaraj
We initiate the study of the inherent tradeoffs between the size of a neural network and its robustness, as measured by its Lipschitz constant. We make a precise conjecture that, f…
Least Squares Regression with Markovian Data: Fundamental Limits and Algorithms
Guy Bresler, Prateek Jain, Dheeraj Nagaraj +2
We study the problem of least squares linear regression where the data-points are dependent and are sampled from a Markov chain. We establish sharp information theoretic minimax lo…
A Corrective View of Neural Networks: Representation, Memorization and Learning
Guy Bresler, Dheeraj Nagaraj
We develop a corrective mechanism for neural network approximation: the total available non-linear units are divided into multiple groups and the first group approximates the funct…