INDEX

Explanations

measurement relative to other things

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

 jestli

-1.84

was

-1.79

 also

-1.77

 perfons

-1.77

</h1>

-1.75

 soms

-1.73

夙

-1.73

⮜

-1.73

]}$

-1.70

 citroen

-1.69

POSITIVE LOGITS

2.97

2.91

the

2.77

2.72

2.55

2.50

 anunció

2.41

2.23

2.22

 dibanding

2.17

Activations Density 0.013%