Curious, since this is a strong loop old frame + input -> new frame, What happens if a non-CS image is used to start it off? Or a map the model has never seen. Will the model play ball, or will it drift back to known CS maps?
468 karma · joined July 14, 2014
Curious, since this is a strong loop old frame + input -> new frame, What happens if a non-CS image is used to start it off? Or a map the model has never seen. Will the model play ball, or will it drift back to known CS maps?
Would be a great step in spreading knowledge across cultures and backgrounds.
It seems unlikely to actually break regulations if it's clear that the data has been fished out of the entropy well.
I want to discuss though if "real" 3D is the right solution for such constrained platforms. The aliasing is quite extreme due to large pixels and lack of filtering. A more constrained 3D renderer or faking 3D with billboards may be better suited for these platforms. Then the aliasing can make it look more like pixel art than the unstructured noise a real render gives.
The high likelihood of collisions is explained by the birthday paradox.
With a 128-bit hash you don't have this problem anymore.
This made me smile. My brain autocompleted the fix to something like "help the neighbors trim their tree", but of course the fix is new hardware.
You can then test a text to see if it increases or decreases the IQ of a model, if included in the training set.
This makes it possible to classify text as "decreases intelligence" of "increases intelligence".
Logical next step is to mark books based on the expected effect on intelligence.
I've had success discussing floats with compute engineers by framing it like a really large fixed point number with all zeroes, except a small bit range. The location and value of the bit range are set by the exponent and mantissa, respectively. This makes precision issues and things like denormals straight forward to discuss.
Just use int64.