INDEX

Explanations

foreign language codes and names

The neuron activates on two‐letter uppercase language or region codes (e.g. “zh,” “FR,” “HU,” “EN”).

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

our

-1.81

it

-1.70

 your

-1.63

 like

-1.38

おわりに

-1.33

 those

-1.27

 that

-1.20

for

-1.13

 stav

-1.09

ススメ

-1.09

POSITIVE LOGITS

''}

1.26

そして

1.23

 reguli

1.19

adering

1.07

 gången

1.05

特别是

1.05

atiles

1.05

 '+':

1.04

""}

1.03

нциклопе

1.02

Activations Density 0.002%