Batel Buaron^1,2, Shlomit Ben-Ami^1,3,4, Roy Mukamel^1,2; ¹Sagol School of Neuroscience, Tel-Aviv University, Tel-Aviv, Israel, ²School of Psychological Sciences, Tel-Aviv University, Tel-Aviv, Israel, ³Minducate science of learning research and innovation center, Tel-Aviv, Israel, ⁴MIT Department of Brain and Cognitive Sciences, Cambridge, MA, USA

To successfully interact with the environment, it is essential to discriminate external events ‎from those that are a result of our actions. Indeed, voluntary actions ‎modulate neural ‎responses to their sensory consequences, relative to responses evoked by identical stimuli ‎from ‎an external source. One potential explanation for such modulations is specific ‎expectation about upcoming events, that may differ between triggering an event vs. ‎observing the same event passively. Another explanation could be driven by a general ‎difference in signal gain between stimulus generation and passive observation, such as ‎encoding agentic source of the stimulus (self/other). To ‎this end, we used an EEG paradigm ‎in which participants learn the ‎coupling between a cue (button press/tone) and a specific ‎visual stimulus (pictures of faces or objects). Results indicate that participants (n=30) successfully ‎learnt the mapping between cues and outcome well across conditions (accuracy above 90%). ‎At the neural level, we examined how ‎learning of new cue-visual contingencies modifies ‎visual evoked responses. We found that P100 amplitudes in the visuo-motor condition were ‎smaller than in the audiovisual condition. Notably, we see a difference in amplitude between ‎cue conditions even on the first repetition of cue-outcome presentation, when no specific ‎outcome expectation has been formed. In addition, we find an effect of learning such that ‎the magnitude of the P100 decreases across repetitions, as mapping between presses/tones ‎and pictures strengthens. Our results suggest that visual evoked responses contain an ‎expectancy component that is not specific to motor commands, in addition to a motor ‎component that is invariant to the degree of mapping between actions and expected visual ‎outcome. Our results provide critical information constraining neural models of visuo-‎motor interactions.‎