3 ms·

Hmm, is faster hallucination better than slower hallucination?

• henry2023 2y ago
  Maybe if you hallucinate fast enough no one will notice?
  • dhosek 2y ago
    I was initially reading it as LLVM performance and thought, “how cool” and then I realized the V was missing and stopped caring.
• ben_w 2y ago
  When the answers can be checked by another program and best-of-n results get better with increasing n, yes. I think TDD code development is this, at least in principle.
  • weebull 2y ago
    If you can write a program to check results you've already solved the issue.
    • ben_w 2y ago
      Sometimes. Some things are easier to verify than to generate. LLMs can often make a distribution of generated outputs that is much closer to a good answer than other methods.