INDEX

Explanations

exclamations and thanks

This neuron lights up on emphatic or polite interjections—tokens carrying strong exclamation or gratitude (e.g. “Alright?!,” “thanks,” “thank you”).

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

</strong>

-1.50

siamo

-1.39

are

-1.32

 オール

-1.28

syd

-1.23

cisco

-1.21

痦

-1.21

 giugno

-1.21

liun

-1.18

POSITIVE LOGITS

totta

1.59

volves

1.49

foreignKey

1.33

htiö

1.33

 هناك

1.30

 любые

1.30

 cuando

1.28

īga

1.24

Some

1.23

 cinéma

1.23

Activations Density 0.009%