DPrivBench: Benchmarking LLMs' Reasoning for Differential Privacy [pdf]
NeurIPS 2026 (Evaluations & Datasets)
DPrivBench: Benchmarking LLMs' Reasoning for Differential Privacy [pdf]
NeurIPS 2026 (Evaluations & Datasets)
Can We Infer Confidential Properties of Training Data from LLMs? [pdf]
NeurIPS 2025 (Spotlight 3%)
Conditional differential measurement error: partial identifiability and estimation [pdf]
NeurIPS workshop CML4Impact 2022
InferPriv: Evaluating Inferential Privacy Leakage in LLM Assistants [pdf]
Preprint 2026
Alignment Defends LLMs from Property Inference Attacks [pdf]
Preprint 2026
Dataset Watermarking with Provable Black-Box Detection [pdf]
Preprint 2026