6 papers
Harnessing Textual Refusal Directions for Multimodal Safety
Moreno D'IncÃ, Nicu Sebe, Massimiliano Mancini
To improve safety in Large Language Models (LLMs) we can either perform post-training alignment or exploit refusal directions in the activation space. Both strategies are less feas…
Metonymy in vision models undermines attention-based interpretability
Ananthu Aniraj, Cassio F. Dantas, Dino Ienco +2
Part-based reasoning is a classical strategy to make a computer vision model directly focus on the object parts that are relevant to the downstream task. In the context of deep lea…
Safe Vision-Language Models via Unsafe Weights Manipulation
Moreno D'IncÃ, Elia Peruzzo, Xingqian Xu +3
Vision-language models (VLMs) often inherit the biases and unsafe associations present within their large-scale training dataset. While recent approaches mitigate unsafe behaviors,…
3D Part Segmentation via Geometric Aggregation of 2D Visual Features
Marco Garosi, Riccardo Tedoldi, Davide Boscaini +3
Supervised 3D part segmentation models are tailored for a fixed set of objects and parts, limiting their transferability to open-set, real-world scenarios. Recent works have explor…
GradBias: Unveiling Word Influence on Bias in Text-to-Image Generative Models
Moreno D'IncÃ, Elia Peruzzo, Massimiliano Mancini +3
Recent progress in Text-to-Image (T2I) generative models has enabled high-quality image generation. As performance and accessibility increase, these models are gaining significant…
OpenBias: Open-set Bias Detection in Text-to-Image Generative Models
Moreno D'IncÃ, Elia Peruzzo, Massimiliano Mancini +6
Text-to-image generative models are becoming increasingly popular and accessible to the general public. As these models see large-scale deployments, it is necessary to deeply inves…