Benchmarking correctly is hard. Criterion generates HTML reports: http://t.co/J15W2yTr2q and performs linear regression to measure noise!
miniblog.
Related Posts
I'm amazed to learn that there is now a marketplace for buying and selling good prompts to midjourney/ChatGPT etc!
You pay and you receive a prompt that generates good results. For example:
Generated Code Generates Overconfident Coders: https://www.deeplearning.ai/the-batch/issue-180/
A study of programmers found that using a LLM for completion produced buggier code but users were more confident in it.
I wonder if this generalises to other completion techniques?
https://twitodon.com/ is a neat tool for finding Mastodon people that you were already following on Twitter. It generates a CSV that Mastodon can consume.
Of the 3,154 folks I follow, it found 117 Mastodon accounts. AIUI it requires other users to use the service.
