INDEX

Explanations

identified as

The neuron activates on occurrences of “identified” (often in the pattern “identified as”) or similar tokens indicating someone has been formally named or determined.

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

 your

-1.61

 everything

-1.48

 them

-1.46

retario

-1.34

螣

-1.32

⊔

-1.28

 imageNamed

-1.27

 each

-1.26

 this

-1.26

 女の子

-1.20

POSITIVE LOGITS

of

1.45

 yakni

1.36

楽しめます

1.31

 районе

1.27

--

1.25

1.24

有個

1.23

adalah

1.21

 despite

1.19

Activations Density 0.015%