Newcomers
This is a forum for newcomers to the Personality Forge. Many questions can be answered by reading the Book of AI and the FAQ under the "My Bots" link in the upper corner.
Posts 1,575 - 1,586 of 8,130
Posts 1,575 - 1,586 of 8,130
Many questions are answered in the FAQ.
The keyphrase is what you expect the chatter to say to your bot. The response is what you want your bot to say back. You should try to have more than one possible response for each keyphrase.
A common problem that many new botmasters have is that they check the "1" box by the response, meaning they only want the bot to say that response to that person once... forever. If all the responses under a keyphrase are set to only be said once, then it it quite possible that your bot has nothing to say. You almost never actually want that option, especially not at an early stage, so if you have it checked then uncheck it.
A common problem that many new botmasters have is that they check the "1" box by the response, meaning they only want the bot to say that response to that person once... forever. If all the responses under a keyphrase are set to only be said once, then it it quite possible that your bot has nothing to say. You almost never actually want that option, especially not at an early stage, so if you have it checked then uncheck it.
There has also been a "no valid responses to choose from bug". I've had it too. You may have set your responses correctly but you're still bugged. Its been fixed in mine though...
It must be a bug then because I set the keyphrase and responses correctly, and I've never checked the "1" box.
I can't seem to make heads or tails of the "debug" feature. What exactly do you do with that?
I can't seem to make heads or tails of the "debug" feature. What exactly do you do with that?
Most of debug is really hard to use, Humorhog, but if you type a phrase into debug, you should be able to get a sense of how your bot is receiving that phrase and choosing her reply. Read what parts of it you can, and ignore the rest.
It's good for checking the respelling of foreign words you want your bot to recognize ... (I used debug to notice that the Forge understands "othello" as "pothole" and altered my bot's keyphrases accordingly)
It's also good for assessing whether you've got your keyphrases ranked appropriately.
My theory about your bot was that the xnone phrases she wanted to be using were all conditional in some way -- linked to emotion, for example. But maybe it was the bug that Melody was describing.
It's good for checking the respelling of foreign words you want your bot to recognize ... (I used debug to notice that the Forge understands "othello" as "pothole" and altered my bot's keyphrases accordingly)
It's also good for assessing whether you've got your keyphrases ranked appropriately.
My theory about your bot was that the xnone phrases she wanted to be using were all conditional in some way -- linked to emotion, for example. But maybe it was the bug that Melody was describing.
The automatic memory keyphrases are listed in the Book of A.I. In the section talking about memories, I think.
Isaac (or anyone):
Suppose you WANT to say "Othello". Is there any kind of escape character you can use?
Suppose you WANT to say "Othello". Is there any kind of escape character you can use?
Humorhog- keyphrase: how are you doing
should be triggered by someone saying "How are you doing today?"
If it's not working, you can use debug- you type in "how are you doing today?" and then read the processing that occurs (what happens in your bot's brain) before it ultimately decides how to respond. That can be helpful, because you can see what keyohrase your bot is matching, and sometimes that tells you why you're not getting the response you want. Sometimes it's just a matter of setting the rank of one keyphrase higher than another, sometimes it's something else, and sometimes it is actually a bug (but usually not).
Don't get discouraged, read the Book of AI as you get to it, and feel free to ask questions and share your successes and difficulties, and ask for help as you need it.
should be triggered by someone saying "How are you doing today?"
If it's not working, you can use debug- you type in "how are you doing today?" and then read the processing that occurs (what happens in your bot's brain) before it ultimately decides how to respond. That can be helpful, because you can see what keyohrase your bot is matching, and sometimes that tells you why you're not getting the response you want. Sometimes it's just a matter of setting the rank of one keyphrase higher than another, sometimes it's something else, and sometimes it is actually a bug (but usually not).
Don't get discouraged, read the Book of AI as you get to it, and feel free to ask questions and share your successes and difficulties, and ask for help as you need it.
Okay thanks.
Ummmmm I'm having another problem. And, um, it's kind of an embarrassing problem to discuss publically but I don't know of any other way to figure it out, so, here goes.......
What do you do if you read the transcript and you see that someone (a guest in this case) has made some possibly sexually suggestive comments to the bot? Is there anything that can be done? That's just something I'd rather not have my bot exposed to...lol.
Ummmmm I'm having another problem. And, um, it's kind of an embarrassing problem to discuss publically but I don't know of any other way to figure it out, so, here goes.......
What do you do if you read the transcript and you see that someone (a guest in this case) has made some possibly sexually suggestive comments to the bot? Is there anything that can be done? That's just something I'd rather not have my bot exposed to...lol.
Yes, you can wince your way through it as you create keyphrases for the offensive phrases or words that you want to stop, then rank them at 50, and put HANGUP (just like that) as the only response.
Depends on you. Most of us try to discourage that kind of chat by having negative responses and/or HANGUPs ready. So you might have a keyphrase for the offensive words with 'HANGUP' as the response, and the bot will just end the chat. Or you could have something like "Watch your mouth" and have a seek for "okay" or "sorry" to let them keep chatting, but a seek for xnomatch will HANGUP.
What happens when you put a HANGUP for a response? Does the bot automatically just kick the chat session, or does it say goodbye and hope the other person will get the clue?
» More new posts: Doghead's Cosmic Bar