Yibo Jiang

Beyond Reward Hacking: Causal Rewards for Large Language Model Alignment

On the Significance of Softmax Geometry: Interpretability and Token Decoding

The Illusion of Role Separation: Hidden Shortcuts in LLM Role Learning (and How to Fix Them)

The Geometry of Categorical and Hierarchical Concepts in Large Language Models

Quantifying Generalization Complexity for Large Language Models

PFT: Enhancing Prompt Injection Robustness via Position-Enhanced Finetuning

Direct Acquisition Optimization for Low-Budget Active Learning