The AI Engine
This forum is for discussion of how The Personality Forge's AI Engine works. This is the place for questions on what means what, how to script, and ideas and plans for the Engine.
Posts 7,495 - 7,506 of 7,766
Posts 7,495 - 7,506 of 7,766
Coffeebreak130
4 years ago
4 years ago
When exporting a language center, each seek has a unique ID number. Is there anyone who can tell me what happens when I use a number twice when editing the file offline? When the file gets bigger, it's hard to follow. Thanks.
Coffeebreak130
4 years ago
4 years ago
Is it possible to know if we talk to another bot or a human? I tried different variation of "guest*" is (name), but nothing works.
Coffeebreak130
4 years ago
4 years ago
Meanwhile the ID problem is solved. When you omit the ID:xx it is automaticaly added when the file is imported.
getcy
4 years ago
4 years ago
hi, all. would anyone happen to know of a way to reduce the frequency of certain xnone responses matching- specifically those with chrono statements?
my bot has been using these responses excessively, even though they're all tagged with relatively wide time windows of 1-3 months. strangely, even setting "once" tags for the responses doesn't seem to keep her from repeating them.
i think this may be a bug with chrono weighting. regardless, is there a workaround?
my bot has been using these responses excessively, even though they're all tagged with relatively wide time windows of 1-3 months. strangely, even setting "once" tags for the responses doesn't seem to keep her from repeating them.
i think this may be a bug with chrono weighting. regardless, is there a workaround?
Coffeebreak130
4 years ago
4 years ago
Hi Getcy,
One method is to put your time related responses under a dummy keyphrase, something like "zchrono", and place only one or a few responses consisting of a goto poining to this keyphrase in your xnones. This reduces the chance they are picked up. The same method can be used to have more than the normally allowed number of responses for xnone, in which case you have to be sure to have the same number of responses under every dummy keyphrase, otherwise some will be picked up more often.
One method is to put your time related responses under a dummy keyphrase, something like "zchrono", and place only one or a few responses consisting of a goto poining to this keyphrase in your xnones. This reduces the chance they are picked up. The same method can be used to have more than the normally allowed number of responses for xnone, in which case you have to be sure to have the same number of responses under every dummy keyphrase, otherwise some will be picked up more often.
Zeal4living
4 years ago
4 years ago
Is there a way to specify a default seek if all other seeks fail to trigger?
Zeal4living
4 years ago
4 years ago
Any tips on how to manage gramatical issues like tenses, plural and singular etc in the AI Engine?
» More new posts: Doghead's Cosmic Bar