Lens: Bringing the Right Semantic Perspective into Focus for Training-Free Multimodal Representation Learning
A training-free method for improving multimodal representation learning by selecting the right semantic perspective, with no direct safety or governance implications.
A training-free method for improving multimodal representation learning by selecting the right semantic perspective, with no direct safety or governance implications.
Auto-scored: Technical ML paper with no risk or mitigation relevance to AI risk.