INDEX

Explanations

person performing an action

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

1.72

of

1.71

not

1.69

’

1.53

 bajos

1.45

in

1.37

違う

1.36

 like

1.32

 period

1.31

量的

1.30

POSITIVE LOGITS

 استقبال

2.14

cheerful

2.04

<unused956>

2.01

сли

1.97

Verifier

1.96

 Thing

1.95

 cheery

1.94

 Kommission

1.93

eyed

1.92

<unused373>

1.91

Activations Density 0.001%