INDEX

Explanations

names and conjunctions

The neuron activates on the names of music (record) labels.

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

 molesta

-1.64

 byen

-1.63

 TAMBÉM

-1.61

籏

-1.56

 felpa

-1.55

の人に

-1.54

 sirena

-1.52

 beneficiario

-1.51

 falda

-1.51

 apartado

-1.47

POSITIVE LOGITS

and

2.36

 which

2.03

 both

1.92

or

1.81

 these

1.73

all

1.72

 several

1.70

to

1.69

 those

1.66

 многие

1.57

Activations Density 0.076%