Finding Visual Task Vectors
This work addresses the problem of reducing reliance on examples for visual tasks, but it is incremental as it builds on existing Visual Prompting techniques.
The authors identified task vectors in a Visual Prompting model that encode task-specific information, enabling the model to perform tasks without input-output examples, achieving better performance than the original model.
Visual Prompting is a technique for teaching models to perform a visual task via in-context examples, without any additional training. In this work, we analyze the activations of MAE-VQGAN, a recent Visual Prompting model, and find task vectors, activations that encode task-specific information. Equipped with this insight, we demonstrate that it is possible to identify the task vectors and use them to guide the network towards performing different tasks without providing any input-output examples. To find task vectors, we compute the average intermediate activations per task and use the REINFORCE algorithm to search for the subset of task vectors. The resulting task vectors guide the model towards performing a task better than the original model without the need for input-output examples.