Comment by ares623 3 months ago Wouldn’t the training or whatever make that unicode sequence effectively a smiley face? 4 comments ares623 Reply jncfhnb 3 months ago Yes, but the same face gets represented by many unique strings. Strings which may more may not be tokenized into a single clean “smiley face” token. scotty79 3 months ago Don't ask ChatGPT about seahorse emoji. astrange 3 months ago That's caused by the sampler and chatbot UI not being part of the LLM. It doesn't get to see its own output before it's sent out. tensor 3 months ago Don't ask humans either, apparently.
jncfhnb 3 months ago Yes, but the same face gets represented by many unique strings. Strings which may more may not be tokenized into a single clean “smiley face” token.
scotty79 3 months ago Don't ask ChatGPT about seahorse emoji. astrange 3 months ago That's caused by the sampler and chatbot UI not being part of the LLM. It doesn't get to see its own output before it's sent out. tensor 3 months ago Don't ask humans either, apparently.
astrange 3 months ago That's caused by the sampler and chatbot UI not being part of the LLM. It doesn't get to see its own output before it's sent out.
Yes, but the same face gets represented by many unique strings. Strings which may more may not be tokenized into a single clean “smiley face” token.
Don't ask ChatGPT about seahorse emoji.
That's caused by the sampler and chatbot UI not being part of the LLM. It doesn't get to see its own output before it's sent out.
Don't ask humans either, apparently.