Related papers: Not All Attention Heads Are What You Need: Refinin…
We investigate the CLIP image encoder by analyzing how individual model components affect the final representation. We decompose the image representation as a sum across individual image patches, model layers, and attention heads, and use…
Recent neural models for image captioning usually employ an encoder-decoder framework with an attention mechanism. However, the attention mechanism in such a framework aligns one single (attended) image feature vector to one caption word,…
Multimodal models like CLIP have gained significant attention due to their remarkable zero-shot performance across various tasks. However, studies have revealed that CLIP can inadvertently learn spurious associations between target…
Standard fairness audits of foundation models quantify that a model is biased, but not where inside the network the bias resides. We propose a mechanistic fairness audit that combines projected residual-stream decomposition, zero-shot…
Mechanistic interpretability improves the safety, reliability, and robustness of large AI models. This study examined individual attention heads in vision transformers (ViTs) fine tuned on distorted 2D spectrogram images containing non…
Domain Generalization (DG) aims to learn a model from multiple source domains to achieve satisfactory performance on unseen target domains. Recent works introduce CLIP to DG tasks due to its superior image-text alignment and zeros-shot…
Recent work has explored how individual components of the CLIP-ViT model contribute to the final representation by leveraging the shared image-text representation space of CLIP. These components, such as attention heads and MLPs, have been…
Vision transformer (ViT) expands the success of transformer models from sequential data to images. The model decomposes an image into many smaller patches and arranges them into a sequence. Multi-head self-attentions are then applied to the…
We present a novel technique for interpreting the neurons in CLIP-ResNet by decomposing their contributions to the output into individual computation paths. More specifically, we analyze all pairwise combinations of neurons and the…
Extending CLIP models to semantic segmentation remains challenging due to the misalignment between their image-level pre-training objectives and the pixel-level visual understanding required for dense prediction. While prior efforts have…
Recent Vision-Language Models (VLMs) have demonstrated remarkable multimodal understanding capabilities, yet the redundant visual tokens incur prohibitive computational overhead and degrade inference efficiency. Prior studies typically…
CLIP is one of the most popular foundational models and is heavily used for many vision-language tasks. However, little is known about the inner workings of CLIP. To bridge this gap we propose a study to quantify the interpretability in…
Contrastive Language-Image Pre-training (CLIP) plays an essential role in extracting valuable content information from images across diverse tasks. It aligns textual and visual modalities to comprehend the entire image, including all the…
CLIP is one of the most popular foundation models and is heavily used for many vision-language tasks, yet little is known about its inner workings. As CLIP is increasingly deployed in real-world applications, it is becoming even more…
Contrastive vision-language models excel in zero-shot image recognition but face challenges in few-shot scenarios due to computationally intensive offline fine-tuning using prompt learning, which risks overfitting. To overcome these…
Zero-shot anomaly detection (ZSAD) enables anomaly detection without normal samples from target categories, addressing scenarios where task-specific training data is unavailable. However, existing ZSAD methods either neglect adaptation of…
Despite the success of large-scale pretrained Vision-Language Models (VLMs) especially CLIP in various open-vocabulary tasks, their application to semantic segmentation remains challenging, producing noisy segmentation maps with…
Large Vision-Language Models (VLMs) have achieved remarkable success in multi-modal reasoning, but their inference time efficiency remains a significant challenge due to the memory overhead during decoding, especially when the query and…
The popularity of Contrastive Language-Image Pre-training (CLIP) has propelled its application to diverse downstream vision tasks. To improve its capacity on downstream tasks, few-shot learning has become a widely-adopted technique.…
Understanding the representation shift on Vision Language Models like CLIP under different augmentations provides valuable insights on Mechanistic Interpretability. In this study, we show the shift on CLIP's embeddings on 9 common…