bilevel optimization
An optimization framework containing two levels of optimization tasks where one is nested inside the other. This is relevant for AI systems where decisions at one level affect the constraints or objectives of another.
- A Single-Loop First-Order Algorithm for Linearly Constrained Bilevel Optimization
- A Single-Loop Gradient Algorithm for Pessimistic Bilevel Optimization via Smooth Approximation
- Adaptive Algorithms with Sharp Convergence Rates for Stochastic Hierarchical Optimization
- Adaptive Neighborhood-Constrained Q Learning for Offline Reinforcement Learning
- Beyond Value Functions: Single-Loop Bilevel Optimization under Flatness Conditions
- Bilevel Optimization for Adversarial Learning Problems: Sharpness, Generation, and Beyond
- Compress Large Language Models via Collaboration Between Learning and Matrix Approximation
- Computational Budget Should Be Considered in Data Selection
- Conditional Gradient Methods with Standard LMO for Stochastic Simple Bilevel Optimization
- Differentiable Generalized Sliced Wasserstein Plans
- Differentially Private Bilevel Optimization: Efficient Algorithms with Near-Optimal Rates
- Learning Theory for Kernel Bilevel Optimization
- On the Complexity of Finding Stationary Points in Nonconvex Simple Bilevel Optimization
- Projecting Assumptions: The Duality Between Sparse Autoencoders and Concept Geometry
- Set Smoothness Unlocks Clarke Hyper-stationarity in Bilevel Optimization
- Stochastic Regret Guarantees for Online Zeroth- and First-Order Bilevel Optimization
- System Prompt Optimization with Meta-Learning
- Taught Well Learned Ill: Towards Distillation-conditional Backdoor Attack
- Unlocker: Disentangle the Deadlock of Learning between Label-noisy and Long-tailed Data