Timezone: »

Text-To-Concept (and Back) via Cross-Model Alignment
Mazda Moayeri · Keivan Rezaei · Maziar Sanjabi · Soheil Feizi

Tue Jul 25 02:00 PM -- 04:30 PM (PDT) @ Exhibit Hall 1 #735

We observe that the mapping between an image's representation in one model to its representation in another can be learned surprisingly well with just a linear layer, even across diverse models. Building on this observation, we propose text-to-concept, where features from a fixed pretrained model are aligned linearly to the CLIP space, so that text embeddings from CLIP's text encoder become directly comparable to the aligned features. With text-to-concept, we convert fixed off-the-shelf vision encoders to surprisingly strong zero-shot classifiers for free, with accuracy at times even surpassing that of CLIP, despite being much smaller models and trained on a small fraction of the data compared to CLIP. We show other immediate use-cases of text-to-concept, like building concept bottleneck models with no concept supervision, diagnosing distribution shifts in terms of human concepts, and retrieving images satisfying a set of text-based constraints. Lastly, we demonstrate the feasibility of concept-to-text, where vectors in a model's feature space are decoded by first aligning to the CLIP before being fed to a GPT-based generative model. Our work suggests existing deep models, with presumably diverse architectures and training, represent input samples relatively similarly, and a two-way communication across model representation spaces and to humans (through language) is viable.

Author Information

Mazda Moayeri (University of Maryland)
Keivan Rezaei (University of Maryland)
Maziar Sanjabi (Meta AI)
Soheil Feizi (University of Maryland)

More from the Same Authors