INDEX

Explanations

options available

The neuron fires on occurrences of the word “available,” marking mentions of options, services, or information being made available.

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

 futuro

-1.63

</i>

-1.48

😗

-1.46

()),

-1.45

 donker

-1.40

😚

-1.39

cosplay

-1.38

 dėl

-1.38

😶

-1.38

GIH

-1.37

POSITIVE LOGITS

↵↵

2.13

 данной

1.89

1.60

 wachs

1.59

 обеспечивает

1.59

you

1.55

 underlines

1.54

 месе

1.52

跟他

1.52

:(

1.52

Activations Density 0.018%