INDEX

Explanations

numerical units digit

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

 dilapidated

-2.98

 ſch

-2.98

 ſame

-2.81

 unſ

-2.59

鬭

-2.58

 bedeckt

-2.52

 gewi

-2.50

 間取り

-2.47

мужские

-2.45

 rangi

-2.41

POSITIVE LOGITS

For

2.89

 know

2.88

 untuk

2.84

 для

2.77

It

2.75

麾

2.59

 voor

2.58

इटम

2.55

</th>

2.53

Activations Density 0.021%