INDEX

Explanations

government actions and initiatives

The neuron detects references to government or regulatory bodies and their actions (e.g., “government,” “regulations,” “initiatives,” “official”).

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

Donkey

-1.45

 любого

-1.45

łość

-1.37

ercicio

-1.30

 outspoken

-1.30

Scénario

-1.30

𝑿

-1.27

Exeunt

-1.27

pijama

-1.27

upaten

-1.27

POSITIVE LOGITS

and

1.76

<b>

1.31

ych

1.12

Same

1.12

IA

1.10

ICA

1.09

んだから

1.09

想不到

1.08

do

1.07

 apparaat

1.04

Activations Density 0.042%