INDEX

Explanations

range() for loops

This neuron never activates—it does not pick up on any pattern or feature.

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

by

-1.67

any

-1.61

 elástico

-1.59

angkap

-1.52

 that

-1.45

諍

-1.40

руются

-1.37

 lädt

-1.37

and

-1.30

 тільки

-1.30

POSITIVE LOGITS

1.91

1.89

of

1.63

1.61

1.49

1.48

1.41

<>

1.27

<strong>

1.25

 lesquelles

1.25

Activations Density 0.001%