INDEX

Explanations

foreign language words

The neuron activates primarily on section headings or title lines (e.g. numbered or all‐caps headings) in the document.

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

 this

-1.55

 today

-1.32

↵

-1.21

几人

-1.15

so

-1.13

ieniem

-1.13

 перио

-1.13

ॲ

-1.11

 извест

-1.11

 який

-1.10

POSITIVE LOGITS

 were

1.53

(!__

1.52

are

1.37

 bawa

1.35

Inconvénients

1.30

 kisah

1.29

Leírás

1.28

llingen

1.27

 gustaMe

1.27

isième

1.27

Activations Density 0.042%