🏢 Pennsylvania State University

Speculative MCTS accelerates AlphaZero training by implementing speculative execution, enabling parallel processing of future moves and reducing latency by up to 5.8x.

This paper introduces a novel offline active learning method for node-level tasks on graphs, incorporating network structure and node covariates to improve efficiency and robustness, especially in noi…

pFedClub: Controllable heterogeneous model aggregation boosts personalized federated learning by generating reasonable-sized, personalized models, significantly cutting computational costs.

Bi-directional Preference Optimization (BiPO) generates superior steering vectors for personalized LLM control, improving upon existing methods by directly influencing the generation probability of hu…

Newton Informed Neural Operator efficiently solves nonlinear PDEs with multiple solutions by learning the Newton solver, enabling faster computation and the discovery of new solutions with limited dat…

Provable near-optimality in meta-RL is achieved using a novel bilevel optimization framework and universal policy adaptation algorithm.

LB-CL: A novel parameter-efficient continual learning method for LLMs that boosts performance and reduces forgetting by leveraging parametric knowledge transfer and maintaining orthogonal low-rank sub…

MERIT-IRL: First in-trajectory IRL framework learns reward & policy incrementally from ongoing trajectories, guaranteeing sub-linear regret.

Decentralized Gradient Descent achieves statistically optimal sparse model learning via implicit regularization, even with communication-efficient truncation.

This paper presents novel algorithms achieving speed-ups in differentially private federated online prediction from experts, addressing both stochastic and oblivious adversaries, with rigorous theoret…

AutoDP automates multi-task learning for joint disease prediction on EHRs, significantly improving performance via automated task grouping and architecture search.