Memory is not an archive, but an active carrying of information into the future. Well organized memory is going to be stored conditionally, with the condition indicating its expected relevance, and thrown away when it expires, such as false beliefs about possible futures.
-
-
-
Replying to @danlistensto
But the correct eviction algorithm is not complicated, is it? And the brain should approximate it!
1 reply 0 retweets 1 like -
Replying to @Plinz
hmmm. qualitatively speaking I can see some characteristics of LRU being analogous to short term memory in the brain. I guess what I was thinking about here was the way that long-term memory is moved into and out of "brain cache" based on conditionals that are pretty opaque.
2 replies 0 retweets 0 likes -
Replying to @danlistensto
What if the division between long term memory and working memory is different than we usually think? I noticed that after 8h anesthesia, patients may return to the moment where the anesthesia took hold without noticing the gap...
1 reply 0 retweets 1 like -
Replying to @Plinz
that's consistent with how I usually think about memory given that I understand anesthesia to sever the input channels that would lead to long term memory formation. the interesting part there is that the conscious monitoring loop resumes with its prior cache pre-warmed
1 reply 0 retweets 0 likes
I used to think that WM is like RAM. If it does not constantly get refreshed by the context, it will evaporate. LTM bakes the information into ROM, so you can access it later. But now I wonder if it can be all modeled as flash? But I still think there are different modes.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.