Zexu Sun

CurES: From Gradient Analysis to Efficient Curriculum Learning for Reasoning LLMs

Prompt and Parameter Co-Optimization for Large Language Models

Solving the Granularity Mismatch: Hierarchical Preference Learning for Long-Horizon LLM Agents

Staying in the Sweet Spot: Responsive Reasoning Evolution via Capability-Adaptive Hint Scaffolding

Generalized Encouragement-Based Instrumental Variables for Counterfactual Regression

Evolving LLMs' Self-Refinement Capability via Synergistic Training-Inference Optimization

Cog-Rethinker: Hierarchical Metacognitive Reinforcement Learning for LLM Reasoning

Invariant Deep Uplift Modeling for Incentive Assignment in Online Marketing via Probability of Necessity and Sufficiency

Uncertainty and Influence aware Reward Model Refinement for Reinforcement Learning from Human Feedback

Learning to Focus: Causal Attention Distillation via Gradient‐Guided Token Pruning

Rethinking Causal Ranking: A Balanced Perspective on Uplift Model Evaluation