Prompt the Unseen: Evaluating Visual-Language Alignment Beyond Supervision
By: Raehyuk Jung, Seungjun Yu, Hyunjung Shim
Potential Business Impact:
Helps computers understand new pictures they haven't seen.
Vision-Language Models (VLMs) combine a vision encoder and a large language model (LLM) through alignment training, showing strong performance on multimodal tasks. A central component in this architecture is the projection layer, which maps visual features into the LLM's embedding space. Despite its importance, its ability to generalize to unseen visual concepts has not been systematically evaluated. To address this, we propose a benchmark for evaluating projection-layer generalization. We adapt object detection datasets (rich in fine-grained annotations) into a prompting format and design train/test splits with disjoint label sets, enabling precise control over seen and unseen concept separation. Experimental results show that the projection layer retains about 79 to 88 percent of the performance on unseen classes compared to seen ones across various settings, suggesting a non-trivial level of generalization even without explicit alignment supervision on those concepts. We further analyze this behavior through a mechanistic interpretability lens. Our findings indicate that the feed-forward network in the projection layer functions like a key-value memory, processing seen and unseen tokens in similar ways. This study introduces a new evaluation framework for alignment generalization and highlights the potential for efficient VLM training with limited aligned data.
Similar Papers
Image Recognition with Vision and Language Embeddings of VLMs
CV and Pattern Recognition
Helps computers understand pictures better with words or just sight.
Hidden in plain sight: VLMs overlook their visual representations
CV and Pattern Recognition
Makes computers better at understanding pictures.
Vision-Language Model for Object Detection and Segmentation: A Review and Evaluation
CV and Pattern Recognition
Helps computers understand images better for many tasks.