AIBullisharXiv – CS AI · Mar 267/10
🧠
SCoOP: Semantic Consistent Opinion Pooling for Uncertainty Quantification in Multiple Vision-Language Model Systems
Researchers developed SCoOP, a training-free framework that combines multiple Vision-Language Models to improve uncertainty quantification and reduce hallucinations in AI systems. The method achieves 10-13% better hallucination detection performance compared to existing approaches while adding only microsecond-level overhead to processing time.