AI RESEARCH
LVLM-Aided Alignment of Task-Specific Vision Models
arXiv CS.CV
•
ArXi:2512.21985v2 Announce Type: replace In high-stakes domains, small task-specific vision models are crucial due to their low computational requirements and the availability of numerous methods to explain their results. However, these explanations often reveal that the models do not align well with human domain knowledge, relying instead on spurious correlations. This might result in brittle behavior once deployed in the real-world. To address this issue, we