4 papers
ShieldCLIP: Selective Safety Alignment for Harmful Content Mitigation in Multimodal Foundation Models
Tobia Poppi, Silvia Cappelletti, Samuele Poppi +4
Multimodal encoders such as CLIP underlie many downstream systems, but their web-scale training data embed harmful associations that safety alignment must suppress without unnecess…
DocAttriBench: Benchmarking Answer Grounding in Document Visual Question Answering
Luca De Grandis, Silvia Cappelletti, William Raccagni +3
Answer grounding in document visual question answering remains an open challenge: most benchmarks lack grounding annotations or provide limited-quality labels, while constructing g…
Do Models Share Safety Representations? Cross-Model Steering for Safe Visual Generation
Tobia Poppi, Silvia Cappelletti, Sara Sarto +5
Recent progress in generative modeling has made safety control a central challenge, yet existing approaches remain largely model-specific, requiring retraining or tailored interven…
Improving LLM First-Token Predictions in Multiple-Choice Question Answering via Output Prefilling
Silvia Cappelletti, Tobia Poppi, Samuele Poppi +5
Large Language Models (LLMs) are increasingly evaluated on multiple-choice question answering (MCQA) tasks using *first-token probability* (FTP), which selects the answer option wh…