Wenlin Chen

Diffusive Gibbs Sampling

It HAS to be Subjective: Human Annotator Simulation via Zero-shot Density Estimation

Leveraging Task Structures for Improved Identifiability in Neural Network Representations

Neural Characteristic Activation Value Analysis for Improved ReLU Network Feature Learning

Meta-learning Adaptive Deep Kernel Gaussian Processes for Molecular Property Prediction

Optimal Client Sampling for Federated Learning

An Evaluation Framework for the Objective Functions of De Novo Drug Design Benchmarks

Causal Representation Learning for Latent Space Optimization