INDEX

Explanations

work and pastoral care

The neuron fires on tokens that denote “work” (or equivalent terms in other languages), i.e. occurrences of the word “work” (and its translations such as Chinese “工作”).

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

 gewerb

-1.41

 mulighed

-1.30

ご相談

-1.30

irsty

-1.27

&#

-1.23

臜

-1.22

}^{*

-1.21

 various

-1.20

 auß

-1.20

 Bezirk

-1.20

POSITIVE LOGITS

ſten

1.48

 ministra

1.44

to

1.32

蕻

1.30

工作

1.23

 Semua

1.22

充分

1.21

 impedir

1.19

 рестора

1.18

lawesi

1.17

Activations Density 0.014%