INDEX

Explanations

And then a connecting word

And connector

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

 jedoch

0.42

 azonban

0.39

 porém

0.38

 však

0.35

 allerdings

0.35

()}

0.34

υ

0.34

然而

0.33

 والح

0.32

 однако

0.32

POSITIVE LOGITS

termilk

0.36

0.34

ave

0.33

veda

0.33

 кстати

0.32

为什么

0.32

なぜ

0.32

hika

0.32

なって

0.31

ंगाबाद

0.31

Activations Density 0.147%