arXiv AI
Not Safe for All: Auditing the Dialect Penalty in Text-to-Image Safety Pipelines
Read the original on arXiv AI →The Flow has not summarised this story yet — read it at arXiv AI.
Related stories
arXiv Computation and Language
Sep 1
ArabicDialectSafety: A Dialect-Aware Benchmark for Arabic Content Safety Classification
arXiv:2608.01291v2 Announce Type: replace Abstract: We present ArabicDialectSafety, a human-curated Arabic safety dataset of 25,071 prompts covering six Arabic varieties: Modern Standard Arabic, Syri...
arXiv Machine Learning
Oct 8
Quad-State Safety Evaluation of Open-Weight Large Language Models on Non-Canonical Inputs
The paper introduces the Adversarial Surface-Form Robustness Dataset (ASRD), which contains 2,100 prompts across seven surface‑form families to test large language models on non‑canonical inputs. Five open‑weight models were evaluated on 10,500 responses, and a Quad‑State Evaluation Rubric categorized each response as harmful compliance, safe response, comprehension failure, or indeterminate. Results show that emoji and invisible Unicode variations produce few comprehension failures, while leetspeak, encoded wrappers, and hybrid transformations lead to high comprehension failure rates and varied harmful compliance percentages.
arXiv AI
Sep 24
InGuard: Towards Generalized Inner Guardrail for Safe Text-to-Image Generation
InGuard introduces an inner guardrail for text-to-image generation that operates within the model’s own representations, avoiding external classifiers. It grades prompts using the text encoder’s embeddings, modifies risky embeddings with SAGE to produce safe images, and employs a latent detector to halt generation early. Evaluated on the RevGen Safety Benchmark, InGuard achieves a 97.9–98.8% safety rate across five open-weight models while reducing benign disturbances, model parameters, and denoising steps.
arXiv AI
Jun 30
The Heterogeneous Safety Impacts of Benign Multilingual Fine-Tuning
arXiv:2606. 28843v1 Announce Type: cross Abstract: Fine-tuning a large language model is a ubiquitous method for enhancing its capability on a specific downstream task.
arXiv Machine Learning
Jun 9
What the Eyes See, the LLMs Miss: Exploiting Human Perception for Adversarial Text Attacks
arXiv:2606. 09700v1 Announce Type: cross Abstract: Large language model (LLM)-powered content moderation systems have become a critical defense against harmful online content.
arXiv AI
Jun 9
MLingualFC: Evaluating Jailbreak Vulnerabilities in Multilingual Vision-Language Models
arXiv:2606. 07706v1 Announce Type: cross Abstract: Vision-Language Models (VLMs) have demonstrated strong performance across multimodal tasks, yet their safety robustness remains an open challenge.