INDEX

Explanations

movement related to rolling

The neuron is activated by occurrences of the root “roll” (and its inflected forms like roll, rolls, rolled, rolling).

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

AfterViewInit

-1.27

BeforeClass

-1.23

even

-1.22

By

-1.20

although

-1.20

``,

-1.18

new

-1.18

aby

-1.17

 لإ

-1.17

llion

-1.16

POSITIVE LOGITS

 เรา

1.46

Tires

1.31

 ktorý

1.31

staa

1.28

啱

1.26

 aplicação

1.24

φέ

1.24

 který

1.24

 doprav

1.21

セッション

1.20

Activations Density 0.057%