Chi Jin

27

Papers

510

Total Citations

Papers (27)

Gradient Descent Can Take Exponential Time to Escape Saddle Points

NeurIPS 2017arXiv

Local Maxima in the Likelihood of Gaussian Mixture Models: Structural Results and Algorithmic Consequences

NeurIPS 2016arXiv

Provable Efficient Online Matrix Completion via Non-convex Stochastic Gradient Descent

NeurIPS 2016arXiv

On the Provable Advantage of Unsupervised Pretraining

Learning World Models for Interactive Video Generation

Tuning-Free Stochastic Optimization

FightLadder: A Benchmark for Competitive Multi-Agent Reinforcement Learning

DOLLAR: Few-Step Video Generation via Distillation and Latent Reward Optimization

Bellman Eluder Dimension: New Rich Classes of RL Problems, and Sample-Efficient Algorithms

Sample-Efficient Learning of Stackelberg Equilibria in General-Sum Games

Sample-Efficient Reinforcement Learning of Partially Observable Markov Games

Efficient Phi-Regret Minimization in Extensive-Form Games via Online Mirror Descent

Optimistic Natural Policy Gradient: a Simple Efficient Policy Optimization Framework for Online RL

DoWG Unleashed: An Efficient Universal Parameter-Free Gradient Descent Method

Is RLHF More Difficult than Standard RL? A Theoretical Perspective

Context-lumpable stochastic bandits

Faster Eigenvector Computation via Shift-and-Invert Preconditioning

Efficient Algorithms for Large-scale Generalized Eigenvector Computation and Canonical Correlation Analysis

No Spurious Local Minima in Nonconvex Low Rank Problems: A Unified Geometric Analysis

How to Escape Saddle Points Efficiently

Is Q-Learning Provably Efficient?

On the Local Minima of the Empirical Risk

Stochastic Cubic Regularization for Fast Nonconvex Optimization

Near-Optimal Reinforcement Learning with Self-Play

On the Theory of Transfer Learning: The Importance of Task Diversity

Provably Efficient Reinforcement Learning with Kernel and Neural Function Approximations

Sample-Efficient Reinforcement Learning of Undercomplete POMDPs