INDEX

Explanations

words ending in -ation, -ated, -ist

The neuron consistently lights up on interior fragments of longer, formal vocabulary words (e.g. acceleration, researched, pursuit, congregationalist, recruits, anticipated).

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

趼

-2.64

It

-2.45

 osoby

-2.33



-2.33

阗

-2.28

郿

-2.25

 otras

-2.25

もあって

-2.17

罴

-2.16

 Preguntas

-2.14

POSITIVE LOGITS

2.56

?"

2.27

2.22

there

2.20

..."

2.19

the

2.19

where

2.09

.........

2.08

:"

2.08

.....

2.08

Activations Density 0.105%