Unmasking Backdoors: An Explainable Defense via Gradient-Attention Anomaly Scoring for Pre-trained Language Models

Unmasking Backdoors: An Explainable Defense via Gradient-Attention Anomaly Scoring for Pre-trained Language Models

5 October 2025

Anindya Sundar Das

ArXiv (abs)PDF HTML Github (191★)

Papers citing "Unmasking Backdoors: An Explainable Defense via Gradient-Attention Anomaly Scoring for Pre-trained Language Models"

0 / 0 papers shown

Title
No papers found