Niloofar Mireshghallah

CIMemories: A Compositional Benchmark For Contextual Integrity In LLMs

Harnessing Optimization Dynamics for Curvature-Informed Model Merging

Operationalizing Data Minimization for Privacy-Preserving LLM Prompting

Spectrum Tuning: Post-Training for Distributional Coverage and In-Context Steerability

Privasis: Synthesizing the Largest "Public" Private Dataset from Scratch

Bob’s Confetti: Phonetic Memorization Attacks in Music and Video Generation

Privacy-Preserving LLM Interaction with Socratic Chain-of-Thought Reasoning and Homomorphically Encrypted Vector Databases

Can Large Language Models Really Recognize Your Name?

Fighting Fire with Fire: Assessing Test Set Contamination Through Deliberate Training on Test Data

Reinforcement Learning Improves Traversal of Hierarchical Knowledge in LLMs

AI as Humanity’s Salieri: Quantifying Linguistic Creativity of Language Models via Systematic Attribution of Machine Text against Web Text

Exploring the limits of strong membership inference attacks on large language models

HAICOSYSTEM: An Ecosystem for Sandboxing Safety Risks in Interactive AI Agents

The Surprising Effectiveness of Membership Inference with Simple N-Gram Coverage

HAICOSYSTEM: An Ecosystem for Sandboxing Safety Risks in Human-AI Interactions

A False Sense of Privacy: Evaluating Textual Data Sanitization Beyond Surface-level Privacy Leakage

ParaPO: Aligning Language Models to Reduce Verbatim Reproduction of Pre-training Data

Leveraging Set Assumption for Membership Inference in Language Models