INDEX

Explanations

predicting follow-up words

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

abad

1.55

of

1.45

اکہ

1.44

ej

1.41

セプト

1.40

ob

1.40

icho

1.39

ele

1.36

obj

1.35

ž

1.30

POSITIVE LOGITS

 dainty

1.90

куса

1.85

 scarred

1.85

餐

1.78

 "/",

1.78

 impressively

1.78

 melanin

1.78

сера

1.77

 căng

1.76

ටි

1.76

Activations Density 0.000%