conftrace_

trigger token

3 papers

Explore in graph

Co-occurring keywords · papers mentioning both

multidimensional analysis (1) orthogonal direction (1) activation variation (1) neuron activation (1) multilingual model (1) representation shift (1) refusal behavior (1) text-to-image diffusion (1) diffusion model (1) backdoor attack (1)

Papers

Char-mander Use mBackdoor! A Study of Cross-lingual Backdoor Attacks in Multilingual LLMs EMNLP 2025 Efficient Input-level Backdoor Defense on Text-to-Image Synthesis via Neuron Activation Variation ICCV 2025 The Hidden Dimensions of LLM Alignment: A Multi-Dimensional Analysis of Orthogonal Safety Directions ICML 2025