INDEX

Explanations

endings of phrases

The neuron selectively lights up on contentful nouns—especially multi‐syllable, topic-defining nouns like “histories,” “philosophies,” “warehouse,” “jokes,” 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

-1.34

left

-1.24

 DOLLARS

-1.23

 anum

-1.20

on

-1.19

 treba

-1.17

\\

-1.16

 Fotografía

-1.15

lihatan

-1.13

 with

-1.12

POSITIVE LOGITS

our

1.56

 считают

1.45

 that

1.42

あぁ

1.29

：“……”

1.29

 zawsze

1.29

 всегда

1.27

 двумя

1.24

誒

1.23

詆

1.23

Activations Density 0.086%