Hacker News new | ask | show | jobs
by TheDudeMan 480 days ago
They do show their model as winning every category in Long Range Arena (LRA) benchmark. Hopefully they have not excluded losing categories or better models.
1 comments
yorwba 480 days ago
Winning against their own baseline, not against the current best-performing model. Which apparently is S5 currently https://paperswithcode.com/sota/long-range-modeling-on-lra with 87.46 overall vs. 58.31 here.
link