Have you tested the latest models with the dataset?
#7
by sssdfa - opened
I want to know how latest models, like GPT-6-ASTRA, Fable-5.1, DeepSeek-v4.1-flash, perform on this dataset.
Hi, thanks for your interest. We haven't planned to do the latest test on those models, but actually we are pretty sure that frontier models could saturate on this benchmark. That's why, on the other hand, our team is on our way of building a new frontier-level physics benchmark. Stay tuned!
StarThomas1002 changed discussion status to closed