INDEX

Explanations

at least

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 многое

1.83

了很多

1.59

 inilah

1.45

Many

1.44

多了

1.43

 stuff

1.36

很多

1.32

すごく

1.31

Much

1.31

了不少

1.28

POSITIVE LOGITS

 mindestens

3.40

当該

2.96

 almeno

2.89

 atleast

2.77

少なくとも

2.61

uktikan

2.59

 Eligible

2.57

 whichever

2.54

 najman

2.51

preferably

2.50

Activations Density 0.890%