INDEX

Explanations

questions and answers/solutions

The neuron is picking out numeric tokens and mathematical expressions (i.e. numbers and equation‐related symbols).

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

 wirklich

-1.58



-1.55

 maailman

-1.48

 THESE

-1.47

谙

-1.46

なければならない

-1.45

 الأخرى

-1.41

 nélk

-1.41

 These

-1.41

UNCIL

-1.38

POSITIVE LOGITS

ではでは

1.44

1.38

1.35

ⅽ

1.27

 before

1.27

さてさて

1.27

and

1.26

HO

1.24

還在

1.21

Activations Density 0.006%