>>7674803
Yeah, no mate. Trying from memory first and then correcting with reference generally yields better learning results. I was pretty sure I'd heard about it in Huberman's podcast, and it turns out the evidence is overwhelmingly supporting of that approach.