In my current implementation of HTM there are no segments within cells. That’s because I don’t know how segments contribute to the model.
Each cell has temporal connections to other cells. The Temporal Pooler checks for active columns and active cells within active columns. For each active cell it checks which temporal connections are connected (permanence > threshold) and if they are, it sets the cell it is connected to to a predictive state.
I don’t think this is the right way. Just a single temporal connection from a currently active cell to another cell shouldn’t be enough to make that cell go in a predictive state. How many temporal connection to a cell from currently active cells should be connected in order for it to become predictive?
Other successful implementations of HTM have also used this optimization (such as etaler for example) @scott commented on this optimization idea on another thread. He mentioned that this is a reasonable optimization and that the algorithm would still work. The cost would be increased likelihood of false positives once you get up to around 10-15 predicted patterns, and he mentioned that this could be mitigated with the right learning rates.
An important aspect is that segments support the reuse of a cell in multiple contexts, without those contexts conflicting and overwriting each other. This ability increases capacity and addresses “catastrophic forgetting”.
This is somewhat subjective to the use-case, but you are correct that in most cases one is not going to be a very good value to use. The setting in vanilla TM is called the activation threshold, and it is primarily associated with balancing capacity and noise tolerance. I did a quick analysis of this property in another thread here.