INDEX

Explanations

taught, conditioned, or indoctrinated

This neuron detects words and phrases related to instruction or conditioning (e.g. taught, to believe, indoctrination, conditioning).

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

for

-1.69

).

-1.63

if

-1.40

-1.39

 posses

-1.38

 동안

-1.38

Arty

-1.31

 "...

-1.30

 decisively

-1.28

ﺀ

-1.27

POSITIVE LOGITS

 culturale

1.37

 giocatore

1.35

 móg

1.34

 ベスト

1.33

gulation

1.30

mathrm

1.27

 takimi

1.26

1.25

utenants

1.25

 detayl

1.25

Activations Density 0.038%