INDEX

Explanations

roll, rolling, rolled, rolls

The neuron flags various forms of the word “roll” (e.g. roll, rolled, rolling, rollover) in financial or account‐rollover contexts.

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

šia

-1.33

less

-1.32

ful

-1.29

坚定

-1.28

as

-1.27

feel

-1.25

риях

-1.24

 there

-1.23

 alguien

-1.23

others

-1.22

POSITIVE LOGITS

nitř

1.34

 into

1.34

 mehrere

1.23

、

1.20

 创建

1.18

 και

1.15

不能

1.13

his

1.12

 sostu

1.11

𖤐

1.11

Activations Density 0.018%