12 papers
TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
Alireza Dehghanpour Farashah, Zhuan Shi, Negar Rostamzadeh +1
Text-to-image diffusion models can be misused to generate harmful content through adversarial or paraphrased prompts that bypass built-in safety mechanisms. Existing concept erasur…
Hidden in Thought: Transferable Chain-of-Thought Artifacts Induce Harmful Behavior
Ali khalil, Aly M. Kassem, Mohamed Abdelrazek +3
We investigate whether harmful chain-of-thought (CoT) traces from compromised language models can transfer unsafe behaviour and be distilled into reusable jailbreak attacks. Using…
Shared Latent Structures Enable Unified Backdoor Detection and Mitigation in LLMs
Omar Mahmoud, Aly M. Kassem, Thommen George Karimpanal +4
Backdoor attacks in large language models (LLMs) are often treated as isolated trigger-response failures, motivating defenses tailored to specific triggers or behaviors. We show th…
A Unified Framework to Quantify Cultural Intelligence of AI
Sunipa Dev, Vinodkumar Prabhakaran, Rutledge Chin Feman +16
As generative AI technologies are increasingly being launched across the globe, assessing their competence to operate in different cultural contexts is exigently becoming a priorit…
Delta-Crosscoder: Robust Crosscoder Model Diffing in Narrow Fine-Tuning Regimes
Aly Kassem, Thomas Jiralerspong, Negar Rostamzadeh +1
Model diffing methods aim to identify how fine-tuning changes a model's internal representations. Crosscoders approach this by learning shared dictionaries of interpretable latent…
Multilingual Amnesia: On the Transferability of Unlearning in Multilingual LLMs
Alireza Dehghanpour Farashah, Aditi Khandelwal, Marylou Fauchard +3
As multilingual large language models become more widely used, ensuring their safety and fairness across diverse linguistic contexts presents unique challenges. While existing rese…