INDEX

Explanations

specific domains and technical terms

This neuron lights up on relatively long, content‐bearing nouns (often multi‐syllable or technical terms) in the text.

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

ícias

-1.40

<>();

-1.30

ў

-1.30

 särskilt

-1.23

について

-1.23

 Fiscalía

-1.23

 egentligen

-1.21

 JavaScript

-1.21

 tillsammans

-1.20

 از

-1.20

POSITIVE LOGITS

imbawa

1.36

but

1.27

 notables

1.21

Життє

1.20

 испы

1.17

it

1.15

 onus

1.14

嚴重

1.13

 Porém

1.12

creme

1.07

Activations Density 0.093%