Methods, systems, and devices for wireless communications are described. In some examples, a transmitting device (e.g., a base station) may utilize machine learning to modify a signal as conditions of a channel change to reduce a peak to average power (PAPR). For example, a base station may select a type of machine learning. The base station may receive one or more feedback messages related to a condition of a channel and modify a downlink signal based on the selected type of machine learning and the one or more feedback messages. In some cases, the base station may transmit the modified downlink signal to a user equipment (UE) along with information indicating the modified downlink signal and the UE may reconstruct the downlink signal based on the information.