Chaoran Chen
2 papers · Latest:
Human-Computer Interaction
Stayin' Aligned Over Time: Towards Longitudinal Human-LLM Alignment via Contextual Reflection and Privacy-Preserving Behavioral Data
A new framework for longitudinal human-LLM alignment reveals that user preferences change over time, challenging static evaluation methods.
2605.04029
Cryptography & SecurityBehavioral Canaries: Auditing Private Retrieved Context Usage in RL Fine-Tuning
Introduces "Behavioral Canaries" to audit if private retrieved contexts are used in RL fine-tuning by detecting stylistic changes, not memorization.
2604.22191
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.