INDEX

Explanations

8 followed by colon

The neuron is strongly activated by numeric tokens (digits and number sequences), e.g. chapter/verse numbers, years, line numbers, etc.

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

-1.55

er

-1.35

,*

-1.33

 cathedral

-1.23

la

-1.22

 tenders

-1.21

-1.20

</h2>

-1.17

re

-1.16

ってみた

-1.16

POSITIVE LOGITS

 juſ

1.51

"}}>

1.44

嚳

1.41

arschijnlijk

1.41

degna

1.35

𝐨

1.35

îner

1.28

vantaged

1.28

];

1.28

 kuitenkin

1.27

Activations Density 0.090%