Feedback-related dynamics of hierarchical error processing in goal-directed actionKroflič, Niko (Avtor) Kunavar, Tjaša (Avtor) De Pauw, Kevin (Avtor) Babič, Jan (Avtor) error-related potentialsfeedback-related negativityperformance monitoringneural signalsbrain-machine interfacesvisuomotor rotationThe performance monitoring system is essential for adaptive behavior and the development of brain-machine interfaces that utilize neural feedback signals. The posterior medial frontal cortex generates different error-related potentials (ErrP), including error-related negativity (ERN), N2, and feedback-related negativity (FRN), which encode specific aspects of performance evaluation. In this study, we reexamine the hierarchical framework of error processing by investigating how low-level execution error detection and correction influence high-level outcome evaluation as reflected in FRN dynamics. Furthermore, we examine whether neural signals associated with outcome errors maintain consistent or distinct feature representations under different experimental conditions of altered feedback availability. Using a visuomotor rotation task, we manipulated the availability of visual feedback in three blocks to examine how immediate sensory error detection and corrective actions interact with outcome processing. Participants (n = 16) performed reaching movements while experiencing unexpected cursor rotations (±20° and ±40°; 20% probability) that challenged their sensorimotor control and task success. EEG recordings revealed that the FRN showed valence sensitivity in Blocks 1 and 3, while Block 2 exhibited a surprise-driven response without outcome differentiation. In contrast, posterior negativity appeared only in Blocks 1 and 3, where participants could detect and correct movement errors. This posterior response emerged on trials requiring corrective movements, regardless of final outcome, and appears to be driven by the availability of sensory feedback and error correction rather than by outcome valence. Furthermore, we demonstrate robust classification between low-level and high-level error signals and their conditional outcome-related variations, providing a foundation for more informative feedback in adaptive neural interfaces.IEEE20262026-04-30 11:18:10Neznano29237UDK: 616.8ISSN pri članku: 1558-0210DOI: 10.1109/TNSRE.2026.3683510COBISS_ID: 275968003ZDAsl© 2026 The Authors.