Anchoring What Matters: A Dual-Level Learning Framework for Visually-Grounded Multimodal Reasoning
A technical paper proposing a dual-level learning framework to improve visually-grounded multimodal reasoning, with no direct safety or risk implications.
A technical paper proposing a dual-level learning framework to improve visually-grounded multimodal reasoning, with no direct safety or risk implications.
Auto-scored: This is a standard academic capability paper with no documented risk or mitigation.