INDEX

Explanations

action verbs

The neuron fires on verbs naming key actions or steps (e.g. produce, protect, transform, describe, build, rule out) typically used to describe processes or methods in technical and scientific text.

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

-1.50

 يؤ

-1.48

by

-1.48

因为它

-1.41

<bos>

-1.40

但是在

-1.37

 there

-1.36

鯖

-1.31

to

-1.30

但却

-1.30

POSITIVE LOGITS

the

1.48

konos

1.41

ονται

1.40

 savait

1.39

our

1.38

 obuv

1.38

 zimowe

1.37

and

1.36

۶

1.34

 sekä

1.33

Activations Density 0.288%