INDEX

Explanations

passenger cars

The neuron fires on occurrences of the token “index” (e.g. in filenames or section headings like “index.md”).

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

-2.02

in

-1.82

on

-1.63

 came

-1.62

 will

-1.59

by

-1.58

 after

-1.56

 After

-1.55

of

-1.50

 became

-1.48

POSITIVE LOGITS

 bénéficier

1.52

 magnifiques

1.50

 láp

1.48

 descuentos

1.45

 probablement

1.43

 véhic

1.43

だいぶ

1.42

 lác

1.38

 doivent

1.38

いきなり

1.36

Activations Density 0.000%