Visual Error Patterns in Multi-Modal AI: A Statistical Approach
Fuente:
arXiv
Gespeichert in:
| 1. Verfasser: | |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866917858686009344 |
|---|---|
| author | Wang, Ching-Yi |
| author_facet | Wang, Ching-Yi |
| contents | Multi-modal large language models (MLLMs), such as GPT-4o, excel at integrating text and visual data but face systematic challenges when interpreting ambiguous or incomplete visual stimuli. This study leverages statistical modeling to analyze the factors driving these errors, using a dataset of geometric stimuli characterized by features like 3D, rotation, and missing face/side. We applied parametric methods, non-parametric methods, and ensemble techniques to predict classification errors, with the non-linear gradient boosting model achieving the highest performance (AUC=0.85) during cross-validation. Feature importance analysis highlighted difficulties in depth perception and reconstructing incomplete structures as key contributors to misclassification. These findings demonstrate the effectiveness of statistical approaches for uncovering limitations in MLLMs and offer actionable insights for enhancing model architectures by integrating contextual reasoning mechanisms. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_00083 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Visual Error Patterns in Multi-Modal AI: A Statistical Approach Wang, Ching-Yi Machine Learning Artificial Intelligence Computer Vision and Pattern Recognition Applications Multi-modal large language models (MLLMs), such as GPT-4o, excel at integrating text and visual data but face systematic challenges when interpreting ambiguous or incomplete visual stimuli. This study leverages statistical modeling to analyze the factors driving these errors, using a dataset of geometric stimuli characterized by features like 3D, rotation, and missing face/side. We applied parametric methods, non-parametric methods, and ensemble techniques to predict classification errors, with the non-linear gradient boosting model achieving the highest performance (AUC=0.85) during cross-validation. Feature importance analysis highlighted difficulties in depth perception and reconstructing incomplete structures as key contributors to misclassification. These findings demonstrate the effectiveness of statistical approaches for uncovering limitations in MLLMs and offer actionable insights for enhancing model architectures by integrating contextual reasoning mechanisms. |
| title | Visual Error Patterns in Multi-Modal AI: A Statistical Approach |
| topic | Machine Learning Artificial Intelligence Computer Vision and Pattern Recognition Applications |
| url | https://arxiv.org/abs/2412.00083 |