I feel like there have been no great leaps in the abilities of llm based tools using llms alone recently. I really doubt having more access to data will actually help this-- most of the "improvements" are using other tools (python, etc) to supplement the llms natural language understanding

Show Post

This is a qq for people who actively work on machine learning systems-- not marketers or stakeholder, not c suite folks-- is there any reason to believe that the current issues with llms and hallucinations can ever be fixed with further training?

Show Post

It’s such a sewer

Show Post

Kapalı içerik

Show Post

Kapalı içerik

Show Post