10 points by killiandunne1 2 days ago | 7 comments
- Can you also measure how often the LLM response makes people laugh? Sometimes the responses that aren't attempting a joke are the funniest, and I'd be more interested in stats of which LLM succeed in that metric.
- Interesting point - thoughts on how to do this? Honestly most models are not-to-kinda funny so I'd be surprised if there were many lol moments. Oral delivery is something I think is v interesting though
- You'll probably still get plenty of smirks, smiles, and silent chuckles. OpenCV or a similar machine vision system should be able to pick those out from a webcam aimed at a reviewer.
- I can’t imagine how to correctly assess the ability to humor, because this is a very subjective and relative phenomenon. The work ahead is serious
- We don't wanna be too serious ;)
- measuring humor might be halfway to measuring taste. congrats on this... really original contribution. wonder if you're planning to evolve the benchmark to incorporate a multi-language dimension. would love to see how Mistral and models built outside the US would perform.
- Nice a good way to 10x inference costs... worth seeing though ahaha
- [dead]