Kong et al.

Reported on 1 benchmark across 1 task

Note: results are matched by exact model name. Different papers may use the same name for different model variants.