Peilin Zhao

PoinnCARE: Hyperbolic Multi-Modal Learning for Enzyme Classification

PoinNSP: Autoregressive 3D Point Cloud Generation with Next-Scale Leve-of-Detail Prediction

Leading the Follower: Learning Persuasive Agents in Social Deduction Games

Curriculum-RLAIF: Curriculum Alignment with Reinforcement Learning from AI Feedback

Silver Stepsize for Faster Zeroth-Order Optimization

Right Question is Already Half the Answer: Fully Unsupervised LLM Reasoning Incentivization

Continual Optimization with Symmetry Teleportation for Multi-Task Learning

Exploring Tradeoffs through Mode Connectivity for Multi-Task Learning

Multi-Task Vehicle Routing Solver via Mixture of Specialized Experts under State-Decomposable MDP

Self-Bootstrapping for Versatile Test-Time Adaptation

COME: Test-time Adaption by Conservatively Minimizing Entropy

Self-Introspective Decoding: Alleviating Hallucinations for Large Vision-Language Models

Scaling Diffusion Language Models via Adaptation from Autoregressive Models

NTPP: Generative Speech Language Modeling for Dual-Channel Spoken Dialogue via Next-Token-Pair Prediction

Efficient Parallel Training Methods for Spiking Neural Networks with Constant Time Complexity

Test-time Adapted Reinforcement Learning with Action Entropy Regularization

TS-LIF: A Temporal Segment Spiking Neuron Network for Time Series Forecasting

Geometric Algebra-Enhanced Bayesian Flow Network for RNA Inverse Design

IgGM: A Generative Model for Functional Antibody and Nanobody Design

Principled Data Selection for Alignment: The Hidden Risks of Difficult Examples

Measuring Diversity in Synthetic Datasets

Evaluating Diversity of LLM-generated Datasets: A Classification Perspective

Heterogeneous Federated Learning: A Dual Matching Dataset Distillation Approach

IgSeek: Fast and Accurate Antibody Design via Structure Retrieval

Parrot: Seamless Spoken Dialogue Interaction with Double-Channel Large Language Models