Methods on performance monitoring for artificial intelligence (ai) / machine learning (ml)

A wireless transmit/receive unit (WTRU) comprising:

• a transceiver;
• and
• a processor configured to:
• receive, via the transceiver, configuration information to determine beam prediction accuracy associated with an artificial intelligence or machine learning (AI/ML) model at the WTRU, wherein the configuration information comprises an indication associated with at least one transmission configuration indication (TCI) state, a first channel state information (CSI) report configuration associated with a first set of reference signal (RS) resources to predict measurements, and a second CSI report configuration associated with monitoring a second set of RS resources;
• based at least on the indication of the at least one TCI state, activate the second set of reference signal (RS) resources to monitor;
• start a performance evaluation counter to determine the beam prediction accuracy;
• in accordance with the first CSI report configuration, predict the measurements for at least one reporting or measurement instance;
• determine the beam prediction accuracy based on the configuration information, the performance evaluation counter, the first set of RS resources, and the second set of RS resources;
• and
• send, via the transceiver, an indication of the beam prediction accuracy based at least on the second CSI report configuration and the performance evaluation counter.