INDEX

Explanations

The neuron detects navigation or pagination markers (e.g. sequences of “>” used to indicate next pages or links).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 appeler

-1.63

.°

-1.46

 hini

-1.45

ྃ

-1.41

 empêcher

-1.41

 descendre

-1.41

넵

-1.41

ishly

-1.38

 manquer

-1.38

FERENCE

-1.38

POSITIVE LOGITS

of

2.06

if

1.77

Though

1.69

 from

1.68

Although

1.59

 that

1.54

 even

1.53

 although

1.52

However

1.52

 Though

1.47

Activations Density 0.013%