INDEX

Explanations

HTML structure tags

The neuron fires on numeric literals (i.e. tokens consisting of numbers or numeric values) in the code/text.

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

 your

-2.72

our

-2.52

 this

-2.05

 their

-1.85

 these

-1.83

釭

-1.77

ſam

-1.77

珰

-1.66

끽

-1.65

ید

-1.64

POSITIVE LOGITS

The

2.22

titial

2.11

 conceivably

1.93

pecabe

1.91

1.81

 albeit

1.77

 diversos

1.75

 montrent

1.75

故意

1.74

 inextrica

1.74

Activations Density 0.003%