6 papers
SEAL: Reinforcing Global Safety in Mixture-of-Experts through Shared Expert ALignment
Qingyu Meng, Yiwei Zha, Jiahuan Pei +3
Mixture-of-Experts (MoE) is a scaling architecture for large language models that activates only a small subset of expert modules per token, enabling massive parameter growth with…
PIIGuard: Mitigating PII Harvesting under Adversarial Sanitization
Mingshuo Liu, Yiwei Zha, Min Chen
Browsing-enabled LLM assistants can fetch webpages and answer contact-seeking queries, creating a practical channel for scraping contact-style personally identifiable information (…
On the Evidentiary Limits of Membership Inference for Copyright Auditing
Murat Bilgehan Ertan, Emirhan Böge, Min Chen +2
As large language models (LLMs) are trained on increasingly opaque corpora, membership inference attacks (MIAs) have been proposed to audit whether copyrighted texts were used duri…
GradEscape: A Gradient-Based Evader Against AI-Generated Text Detectors
Wenlong Meng, Shuguo Fan, Chengkun Wei +5
In this paper, we introduce GradEscape, the first gradient-based evader designed to attack AI-generated text (AIGT) detectors. GradEscape overcomes the undifferentiable computation…
TA3: Testing Against Adversarial Attacks on Machine Learning Models
Yuanzhe Jin, Min Chen
Adversarial attacks are major threats to the deployment of machine learning (ML) models in many applications. Testing ML models against such attacks is becoming an essential step f…
iGAiVA: Integrated Generative AI and Visual Analytics in a Machine Learning Workflow for Text Classification
Yuanzhe Jin, Adrian Carrasco-Revilla, Min Chen
In developing machine learning (ML) models for text classification, one common challenge is that the collected data is often not ideally distributed, especially when new classes ar…