EXPLANATIONS
[传] => [传]
INFERENCE
IFRAME INTEGRATION
NEGATIVE LOGITS
prep-16.000
await-16.000
rio-14.313
brains-14.250
stud-14.188
cat-14.000
his-13.813
outcome-13.813
coin-13.688
prepared-13.625
POSITIVE LOGITS
第三代17.250
_которое16.250
倒闭16.250
_Style16.125
uctor16.000
_Occup15.938
egade15.938
口气15.875
-validator15.750
ometrics15.750
ACTIVATION TIMES 0.0275%
Activation histogram is not available.
Logits histogram is not available.
TOPACTIVATIONS
SUBSAMPLING80%
SUBSAMPLING60%
SUBSAMPLING40%
NONACTIVATING