INDEX

Explanations

followed by a state or action

The neuron specifically detects occurrences of the word “continued.”

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

to

-1.74

fications

-1.61

 mengisi

-1.59

 urusan

-1.58

 gastos

-1.54

Aiheesta

-1.54

 menjalani

-1.52

iors

-1.52

idated

-1.52

 obro

-1.50

POSITIVE LOGITS

1.70

);

1.69

 comprenant

1.68

1.56

1.54

ing

1.50

ra

1.46

ｏ

1.46

and

1.42

\":

1.42

Activations Density 0.012%