INDEX

Explanations

punctuation and surrounding text

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 memeriksa

-1.91

复制代码

-1.89

釭

-1.74

觧

-1.72

 helg

-1.66

 actif

-1.66

 courriel

-1.66

鼉

-1.64

 courrier

-1.63

 mengerti

-1.61

POSITIVE LOGITS

 were

2.00

 serán

1.63

 will

1.62

of

1.52

UTTON

1.49

ят

1.48

1.47

in

1.46



1.41

⽼

1.39

Activations Density 0.002%