Interaction-Grounded Learning with Action-inclusive Feedback
This addresses a limitation in IGL for applications like Brain-computer or Human-computer interfaces, representing an incremental improvement.
The paper tackles the problem of Interaction-Grounded Learning (IGL) failing when feedback includes action information, by developing an algorithm that works with action-inclusive feedback, supported by theoretical guarantees and large-scale experiments.
Consider the problem setting of Interaction-Grounded Learning (IGL), in which a learner's goal is to optimally interact with the environment with no explicit reward to ground its policies. The agent observes a context vector, takes an action, and receives a feedback vector, using this information to effectively optimize a policy with respect to a latent reward function. Prior analyzed approaches fail when the feedback vector contains the action, which significantly limits IGL's success in many potential scenarios such as Brain-computer interface (BCI) or Human-computer interface (HCI) applications. We address this by creating an algorithm and analysis which allows IGL to work even when the feedback vector contains the action, encoded in any fashion. We provide theoretical guarantees and large-scale experiments based on supervised datasets to demonstrate the effectiveness of the new approach.