This is what they wrote:

"For a subset of those trials (N=2,770 trials), 982 images were viewed by all four subjects. Those trials were used as the test dataset, while the remaining trials (N=24,980) were used as the training dataset."

That makes it sound like 982 images were not used for training

> so their MRI->latent space model has seen every one of the 10,000 images in the dataset

Are you sure about that? I wasn't able to understand their test method from the paper, but it sounds like they held out some images from training

No need to be snarky.

If cataracts just makes things look blurry then you're right.

If cataracts makes bright colors look dull, for example, then he would still choose paints that match the true colors of the scene. The bright colors in the scene would look dull to him, but his bright colored paints would look exactly as dull.

That explanation doesn't really make sense, though. His cataracts would affect his vision of his paint and painting just as much as the landscape, so the painting would look similar to the landscape

Thanks. Aside from the weight of the deck itself, it will only need to support a few deck chairs and a few adults. And yes, it will be low profile. The joists will be just an inch off the ground and the surface of the deck around 8 inches high.

Thanks for the input. Yes, it does freeze where I'm at, but I'm thinking frost heave won't be a big deal for a deck that sits a few inches off the ground.

The deck is indeed just for a few deck chairs and doesn't need to be built like a tank. I honestly don't even care if it becomes a little unlevel or wobbly after a few years as long as it continues to support two adults and their beers.

Thanks, that makes sense. Those plans might be shit, but my impression is that all the regulations and suggestions I'm reading are overkill for an unattached deck that sits at ground level and only supports a few patio chairs. Like, do I really need to dig a 4 foot deep hole and dump in bags of concrete for something this size?

Do I need a deck block under every joist? In these plans, they use a deck block under every-other joist:
https://gardenplansfree.com/furniture/deck-plans-free/

but they're using 2x8's, so maybe they don't exceed the max span for 2x8s if the beams on either side count as support?

It will be very low, almost sitting on the ground, so I don't need rails or stairs. I'm trying to place enough blocks to avoid springiness.

Thanks for the suggestions! Unfortunately, I already bought like 20 deck blocks that have a space for a single 2x and therefore can't fit a double beam (this kind https://www.lowes.com/pd/Common-7-in-x-11-in-x-11-in-Actual-7-125-in-x-10-375-in-x-10-375-in-Concrete-Deck-Block/50113084 ). I suppose I could return them all.

Alternatively, could it make sense to use 3 beams like you suggested, but put the blocks under joists instead of beams like in my picture? I got that idea from these plans:
https://gardenplansfree.com/furniture/deck-plans-free/

I can't use 2x8's because I need to keep the deck low to the ground since I'm building it as a "step down" from another deck next to it.

"Your drawing shows some spans are not supported and are 10' long"
I used the strategy from these plans in which only every-other joist has blocks:
https://gardenplansfree.com/furniture/deck-plans-free/

It seems overkill to have block(s) under every joist.

"Your deck blocks should be under your beams (dbl 2x6) and the middle deck blocks would need a beam. Supporting 1 joist midway, but not the next one is not proper."

For both of those things, I took inspiration from these plans:
https://gardenplansfree.com/furniture/deck-plans-free/

which have essentially the same setup for a 12x12' deck with 2x8's.

Are those plans no good, or are my deviations from those plans no good?

I added the image to the original post.

Here is an image of my deck plans to help visualize the setup in question:
https://drive.google.com/file/d/1Zs97wA_hXPqVcCFDbkHZ_RisW3bhXe7K/view?usp=share_link