(A) The generative and inference models of PV-RNN in an MTRNN setting, (B) the error regression graph during tests, and (C) the error regression process. In panels A and B, black lines represent the generative model and red lines show the inference model, with solid red lines showing the feedforward computations of the inference model and dashed red lines showing the BPTT that is used to update in panel A and in panel B. The gray area in panel B represents a two-step temporal window of the immediate past in which is modified to maximize the lower bound. Panel C illustrates the error regression process. At , predictions are generated (left) after observing . The three-time step time window is slid by one time step to (middle; now, ), and an error is observed between the prediction and the target value . The lower bound is computed, and backpropagation is performed; is then optimized, and the prediction is updated (right). This backpropagation/optimization/prediction cycle can be repeated multiple times before moving on to the time window.
This site uses cookies. By continuing to use our website, you are agreeing to our privacy policy. No content on this site may be used to train artificial intelligence systems without permission in writing from the MIT Press.