INDEX

Explanations

achieve, forwards, Ronald, Fed, Lind

target, TARGET, rope, Log

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

上帝

0.35

 ისინი

0.35



0.35

utim

0.33

它可以

0.33

പ്പിച്ച്

0.32

 उसको

0.31

 এরা

0.31

 இறைவன்

0.31

 ЕС

0.31

POSITIVE LOGITS

 hét

0.38

 परिसरात

0.37

 aforementioned

0.36

 обозна

0.36

 kawasan

0.36

koc

0.34

上述

0.34

 нашем

0.34

 chase

0.33

 bijvoorbeeld

0.33

Activations Density 0.285%