INDEX

Explanations

correct usage or form

This neuron detects forward‐slash characters ("/") used as separators.

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

 gange

-1.33

罍

-1.33

画

-1.28

ENING

-1.27

"},

-1.25

 Reſ

-1.23

 urat

-1.23

 finir

-1.23

ᜀ

-1.22

鲠

-1.20

POSITIVE LOGITS

or

1.68

 also

1.60

not

1.55

in

1.38

 insisting

1.23

 often

1.21

 també

1.20

 Also

1.19

It

1.16

}$.

1.14

Activations Density 0.087%