There's an idea of 'fixed action patterns' where certain behaviors are encoded in the neurology of an animal, and triggered by certain stimulus.
For spiders then there is an action pattern that upon certain stimulus they will use their spinnerets begin forming a web, first placing anchor points and gradually building it up with radial threads and spiral. The exact behavior differs by species such that they have different resulting patterns. It seems possible that there is an element of 'learning' beyond the basic pattern as well, even in seemingly very simple animals.
I doubt the spiders have a mental model of what their web looks like as such, it's more evolutionary driven genetic behavior, a particular neuronal pattern encoding the action sequence rather than an underlying fundamental understanding.
For spiders then there is an action pattern that upon certain stimulus they will use their spinnerets begin forming a web, first placing anchor points and gradually building it up with radial threads and spiral. The exact behavior differs by species such that they have different resulting patterns. It seems possible that there is an element of 'learning' beyond the basic pattern as well, even in seemingly very simple animals.
I doubt the spiders have a mental model of what their web looks like as such, it's more evolutionary driven genetic behavior, a particular neuronal pattern encoding the action sequence rather than an underlying fundamental understanding.