aboutsummaryrefslogtreecommitdiff
path: root/include
AgeCommit message (Collapse)Author
45 hoursHash collisionstslil clingman
45 hoursFixed some important bugstslil clingman
45 hoursAttempting Zobrist hashingtslil clingman
45 hoursNot correct usage, need to try hashingtslil clingman
45 hoursno speed increase ...tslil clingman
45 hoursTrying caching to speed things uptslil clingman
45 hoursThanks compile warnings!tslil clingman
45 hoursStarting work on LCD librarytslil clingman
45 hoursFinally (?) fixed winning avoidancetslil clingman
45 hoursRenamingtslil clingman
45 hoursMany fixes, i think this is actually correcttslil clingman
45 hoursExperimentingtslil clingman
45 hoursMerge branch 'pi'tslil clingman
45 hoursCompiler warningtslil clingman
45 hoursFinally fixed move generationtslil clingman
45 hoursCan't stop half-way through turntslil clingman
45 hoursIn-lined as much as a i daretslil clingman
45 hoursFinally fixed memory order of weights!tslil clingman
45 hoursOpps, wrong stonetslil clingman
45 hoursInlined placements, should be fastertslil clingman
45 hoursMemory oob fixtslil clingman
45 hoursRevert "Starting work on MCU port"tslil clingman
This reverts commit 43eec01361eb6174142180c29afef960e041514e.
45 hoursFirst steps towards character LCDtslil clingman
45 hoursLeftover from another branchtslil clingman
45 hoursStarting work on MCU porttslil clingman
45 hoursNew weights from a wider trainingtslil clingman
45 hoursBeginning to work on probabilistic depth now that MCU is outtslil clingman
45 hoursWent back up to 1986 weightstslil
45 hoursQuick sketch of non-determinismtslil
45 hoursAnother road bugtslil
45 hoursDragon is not separate, fixed concession, allowing playing as b/wtslil
45 hoursChanged function back to macrotslil
45 hoursThere was a bug in the minimax!tslil
45 hoursBetter trained weights, an important oversight in minimax (steps <=)tslil
I still can't tell whether if ((ply & 1) == 0) this = this - 1.0; is correct.
45 hoursSo after a small tweak, it actually plays well!tslil
45 hoursSeems tolerable, but not great. Still wont win or lose ???tslil
45 hoursTrying alpha-beta pruning. There's a bug somewhere ...tslil
45 hoursFixed minimax (!), fixed bugs in tak.ctslil
With minimax of depth 1 the evaluation function seems alright with the current method of training and generating weights
45 hoursTrying minimaxtslil
45 hoursCrass errortslil
45 hoursWeights?tslil
45 hoursWider?tslil
45 hoursTrying various things to teach titslil
45 hoursHave to account for CPStslil
45 hoursWell it works, and it's bad. Many bugs fixed in interimtslil
45 hoursClosing in!tslil
45 hoursAuto-generating weightstslil
45 hoursNew weightstslil
45 hoursone output (of course), Pade approximant of logistic for activationtslil
45 hoursTrain on data that is close to the end of the game onlytslil