StrategyQA and Big Bench
Data Skeptic
00:00
The Bleeding Edge Benchmark
Right now the main user of this benchmark is just the big tech companies that can build these large language models. So using all these data sets can be a nice benchmark. The actual paper just came out two months ago, so it's kind of new to the community. It covers mostly like models from Google and OpenAI,. Like GPT and also the recent models Palm; another one I forgot its name.
Transcript
Play full episode
