INDEX

Explanations

numbers and letters like L

The neuron fires on numeric tokens (integers and decimals) in code or data (e.g. coordinate values, dimensions, measurements).

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

 invester

-1.90

 medier

-1.90

‪.‬‬

-1.89

 myndig

-1.81

跸

-1.79



-1.76

 costuras

-1.75

 barcel

-1.74

("")]

-1.73

 måde

-1.72

POSITIVE LOGITS

2.55

like

1.99

来て

1.81

 jednej

1.80

 they

1.73

他的

1.71

 insane

1.70

selbe

1.65

自分が

1.65

 kinda

1.65

Activations Density 0.001%