INDEX

Explanations

do not, will be, here below

The neuron fires strongly on personal, self-referential words—especially pronouns and auxiliaries (I, you, my, me, am, will, can, 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

);

-1.32

 другие

-1.27

-1.23

-1.19

-1.18

 новый

-1.16

“

-1.13

",

-1.13

-1.09

-1.06

POSITIVE LOGITS

 великолеп

1.77

 this

1.71

all

1.66

 even

1.52

 also

1.52

 your

1.50

 each

1.49

Preparazione

1.49

 kesan

1.46

(!__

1.38

Activations Density 0.111%