INDEX

Explanations

ages and counts

The neuron primarily responds to standalone hyphen/dash tokens (e.g. those joining or preceding numbers).

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

for

-1.71

the

-1.66

 each

-1.59

if

-1.54

it

-1.53

 which

-1.47

to

-1.46

 this

-1.45

by

-1.43

痦

-1.39

POSITIVE LOGITS

 computadoras

1.69

૪

1.63

nimmt

1.52

奚

1.47

 freio

1.46

 leichte

1.45

continence

1.43

 bañera

1.41

 {}));

1.41

telah

1.38

Activations Density 0.103%