conftrace_

harmful content mitigation

2 papers

Explore in graph

Co-occurring keywords · papers mentioning both

backdoor removal (1) safety backdoor (1) safety compromise (1) lifecycle mitigation (1) large language model (1) embedding perturbation (1) model customization (1) data curation (1) adversarial defense (1) backdoor attack (1)

Papers

Data to Defense: The Role of Curation in Aligning Large Language Models Against Safety Compromise EMNLP 2025 BEEAR: Embedding-based Adversarial Removal of Safety Backdoors in Instruction-tuned Language Models EMNLP 2024