Qin-Wu Xu

D AMPED G AUSS –N EWTON S EARCH

Reducing Hallucination in Multimodal Large Language Models through Hard Grounding Preference Supervision

Results show that decoder-side adaptation alone can improve a meaningful subset of grounding failures under a fixed multimodal representation, while preserving broader multimodal capabilities.

Variational Template Matching with Statistical Fusion for Anomaly Detection in Patterned Structures

Anomaly detection in structured images is challenging in small-data settings where deep learning approaches are costly or impractical. Classical template matching is simple and interpretable but lacks robustness to geometric variations such as scale, rotation, and perspective. We propose a variational template matching...