ALEXANDRIA, Va., April 21 -- United States Patent no. 12,608,861, issued on April 21, was assigned to HiddenLayer Inc. (Austin, Texas).
"Model output steering using universal images" was invented by Ravikumar Balakrishnan (Beaverton, Ore.) and Mansi Phute (Atlanta).
According to the abstract* released by the U.S. Patent & Trademark Office: "An output of a vision-language model (VLM) can be steered by receiving an input that includes both an image and text, and compositing the image with a universal image configured to elicit specific model activations associated with a targeted behavioral change. The modified image, together with the text, is then supplied to the VLM, resulting in an output that reflects the desired behavioral change. Thi...