However, we may use OOPS to perform bias-optimal proof search for the more recent Gödel machine, which makes provably optimal self-improvements in such general environments.
OOPS can solve tasks unsolvable by traditional reinforcement learners and AI planners, such as Towers of Hanoi with 30 disks (minimal solution size exceeds 10^9).
In our experiments OOPS demonstrates incremental learning by reusing previous solutions to discover a prefix that temporarily rewrites the distribution on its suffixes, such that universal search is accelerated by a factor of 1000.
Alternatively, OOPS can be used to generate informative macros for all residues in your protein, and the macros may even have the same name as the residue.
OOPS will write most of the macros to that directory so these files won't clutter your working O directory (use: "mkdir oops" to create this directory).
When OOPS is finished (and if there were any bad residues), all you have to do is type: "@oops.omac" from within O. This will then automatically centre on the first bad residue and print a message which tells you what is suspicious about this residue.
