Recent searches
Search options
Anthony Shaw
@tonybaloney@fosstodon.orgFirst day back. New blog post! I try and get the world’s smartest AIs to solve a game of Set, debate whether Harry Kane won the Ballon D’or, and test an LLM hack to get it to stop lying so much.
www.linkedin.comCan you prompt LLMs to admit when they don't know the answer?Toward the end of 2024, there were a bunch of releases and announcements for new AI models. In particular, there was a lot of discussion about models which can do “complex reasoning” and “chain of thought” to solve complex problems.