9 papers
Rethinking Neural Network Learning Rates: A Stackelberg Perspective
Sihan Zeng, Sujay Bhatt, Sumitra Ganesh
Neural networks are typically trained with a single learning rate across all layers. While recent empirical evidence suggests that assigning layer-specific learning rates can accel…
A Hessian-Free Actor-Critic Algorithm for Bi-Level Reinforcement Learning with Applications to LLM Fine-Tuning
Sihan Zeng, Sujay Bhatt, Sumitra Ganesh +1
We study a structured bi-level optimization problem where the upper-level objective is a smooth function and the lower-level problem is policy optimization in a Markov decision pro…
Learning in Stackelberg Mean Field Games: A Non-Asymptotic Analysis
Sihan Zeng, Benjamin Patrick Evans, Sujay Bhatt +3
We study policy optimization in Stackelberg mean field games (MFGs), a hierarchical framework for modeling the strategic interaction between a single leader and an infinitely large…
PADME: Procedure Aware DynaMic Execution
Deepeka Garg, Sihan Zeng, Annapoorani L. Narayanan +2
Learning to autonomously execute long-horizon procedures from natural language remains a core challenge for intelligent agents. Free-form instructions such as recipes, scientific p…
Approximate Equivariance in Reinforcement Learning
Jung Yeon Park, Sujay Bhatt, Sihan Zeng +4
Equivariant neural networks have shown great success in reinforcement learning, improving sample efficiency and generalization when there is symmetry in the task. However, in many…
Generating Structured Plan Representation of Procedures with LLMs
Deepeka Garg, Sihan Zeng, Sumitra Ganesh +1
In this paper, we address the challenges of managing Standard Operating Procedures (SOPs), which often suffer from inconsistencies in language, format, and execution, leading to op…