miniblog.

Using machine learning to decide what structure to use for your neural net: https://arxiv.org/pdf/1703.01041.pdf It's pretty meta, but ML is often good at trading computation for results.

A recent release of Free Pascal has a rather interesting behaviour for its list implementation. Lists grow additivitely (rather than multiplicatively) above 128MiB, trading off computational complexity for trying to avoid out-of-memory issues! https://wiki.lazarus.freepascal.org/User_Changes_3.0.4#TList_auto-growth