Every organization differs, of course, making this difficult to usefully answer.
I have always prefered seeing senior devs as capable of architecting and implementing solutions (better yet delegating parts to juniors) such that they can launch a product themselves, where a mid level dev can launch individual features but would struggle with the complexity of choosing both an auth model and a persistence layer for it (mixing well with other persistence uses) and then what the front end should look like overall, and whether to use websockets or... and a junior can fix bugs, or work on tickets/scaffolded tasks in a productive manner.
I have always prefered seeing senior devs as capable of architecting and implementing solutions (better yet delegating parts to juniors) such that they can launch a product themselves, where a mid level dev can launch individual features but would struggle with the complexity of choosing both an auth model and a persistence layer for it (mixing well with other persistence uses) and then what the front end should look like overall, and whether to use websockets or... and a junior can fix bugs, or work on tickets/scaffolded tasks in a productive manner.