INDEX

Explanations

father or mother titles

This neuron detects kinship-relation terms (e.g. “father,” “mother,” “son,” “child”) in titles or descriptions.

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

omorph

-1.47

alties

-1.43

-1.36

 niektor

-1.35

齜

-1.34

 dispuestos

-1.34

 apreci

-1.33

 erwähnten

-1.32

 Geschä

-1.29

 Coloque

-1.29

POSITIVE LOGITS

1.60

1.46

1.34

’

1.31

 كافة

1.30

mgang

1.21

all

1.20

ၜ

1.20

𝗲

1.20

The

1.19

Activations Density 0.017%