Abstract
Implicit adaptation plays an unconscious yet key role in maintaining precise motor control when faced with changes in the body or environment. While models suggest this process is constrained by a hard-wired limit or modulated by error attribution, most insights come from prolonged training, leaving the sensitivity of initial implicit adaptation to error size, type, and timing poorly understood.
To address this, we conducted movement-contingent Single-Trial Learning (STL) adaptation experiments with visuomotor rotations. Participants performed reaching movements under rotated cursor feedback (1°–90°) in trials preceded and succeeded by aligned trials. By manipulating task error (dot vs. arc targets) and feedback timing (terminal vs. delayed), we captured how these variables modulate early adaptation. Additionally, a long-exposure block using a 20° rotation allowed us to compare STL-derived predictions to initial rates of change for prolonged adaptation.
Our findings show initial implicit aftereffects remain stable across perturbation sizes, ramping up between 1° to 15° rotations and remaining at ~6° for larger rotations of 15° to 90°, suggesting a capped response rather than attenuation with error size. Task error removal reduced aftereffects (to ~4°), as did terminal and delayed feedback (to ~2°). STL-derived predictions correlated modestly with early adaptation during prolonged exposure.
These results suggest that initial implicit adaptation is highly sensitive, saturates at a fixed limit, is modulated by task error, feedback timing, and can be meaningfully probed through STL, offering a potential proxy for predicting long-term learning outcomes.
