I find this easiest to practice in a mirror because I have a lot of top-down perceptions of my face and body.

Exercise 1: Point your eyes at the periphery of your face and then look at your face in peripheral vision. I see flickering of different interpretations of my self-image and facial features.

Note: Hallucinations from vision going grey from lack of saccading can be less useful because they mean you've lost bottom-up information.

Exercise 2: Open your eyes at the mirror, without the intention to look at any specific feature in it. Feels similar trying to pay attention to your entire visual field, or trying to pay attention to something in the periphery without saccading.

Practicing exercise 2 seems to let me eventually start saccading around while holding the intention to "not look at anything in particular". Parts of the image keeps disappearing from my conscious vision, similar to how eyesight is just "gone" when you close your eyes and aren't looking at the blackness behind them.

At this point I can do exercise 1 on my whole body at once, while my vision switches between modes:

• Fully integrated
  • Blindsight regions appear and disappear in response to my attention moving around.
  • Blindsight regions often reappear because I noticed they were gone.
• Portions of the visual field having fully integrated interpretations but can't merge at the seams.
• Features split apart each implying a clashing gestalt.

This is really uncomfortable and I'm not sure why I would have something that feels like "opening my eyes without choosing anything to look at anything but still having conscious control of saccades and attention" that feels entirely different from normal vision which doesn't glitch around like this.

What learnings do you draw from this?

It is scary to realize how much of our perception is biased by preconceptions, which I assume is what you call top-down. I once noticed in real time how my perception of a portrait changed tangibly the moment I realized that the face belonged to ”me” (this was before my assumed stream entry, even before starting a daily practice, when I still believed in selves). At first glance it looked like a rather pretty face, but as I saw that it was my face, the proportions changed radically as if it was transformed into a caricature. I can’t even tell which one of the perceptions was most accurate.

A brand new dissertation (in German) by the autistic researcher Hajo Seng discusses the differences between top-down and bottom-up thinking, claiming that autistic people (like me) tend to have bottom-up cognitive processing and reclaming that as something good. I often think about his theories when I reflect upon my meditation.