The logmax part makes it different from something like the Kullback–Leibler divergence, but that might be a good feature—a logmax definition seems harder to hack.
If worrying about butterfly effects and similar, it might be useful to do something like this: let p be the probability distribution of future states given that a burrito is not made, and q the same distribution given that the burrito is made. If p and q are very different (as measured by KL divergence or the approach here) then that means that either a) the burrito is dangerous, or b) the AI can unravel butterfly effects. If p and q are very different for many different burrito it could make, then we have a butterfly effect problem. If they are only different for some burritos, then we have identified the high-impact burritos.
The logmax part makes it different from something like the Kullback–Leibler divergence, but that might be a good feature—a logmax definition seems harder to hack.
If worrying about butterfly effects and similar, it might be useful to do something like this: let p be the probability distribution of future states given that a burrito is not made, and q the same distribution given that the burrito is made. If p and q are very different (as measured by KL divergence or the approach here) then that means that either a) the burrito is dangerous, or b) the AI can unravel butterfly effects. If p and q are very different for many different burrito it could make, then we have a butterfly effect problem. If they are only different for some burritos, then we have identified the high-impact burritos.