INDEX

Explanations

greetings and thanks

This neuron fires strongly on the first (typically capitalized) word of a sentence or comment, e.g. “I,” “Thanks,” “You,” “The,” “Is,” etc.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ACCIÓN

-1.21

to

-1.13

of

-1.09

 zahlreiche

-1.08

 before

-1.05

 until

-1.04

Merkmale

-1.04

㎎

-1.03

 seines

-1.03

or

-1.03

POSITIVE LOGITS

1.27

1.20

 elétrico

1.20

の基本

1.17

1.13

 leçon

1.10

 やすい

1.09

 píše

1.09

 remplacer

1.05

 oznac

1.05

Activations Density 0.090%