INDEX

Explanations

game controller button

This neuron fires on occurrences of the word “New” (in English or its translations/localized forms, often in metadata or locale‐label contexts).

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

-1.41

-1.40

 where

-1.37

-1.35

-1.34

…

-1.33

-1.30

—

-1.29

-1.28

\\

-1.25

POSITIVE LOGITS

 rafraî

1.53

 réfrig

1.45

 refroid

1.42

 côtés

1.41

 یا

1.39

 grotte

1.36

 épar

1.34

 plastiques

1.34

最近は

1.34

หรือ

1.34

Activations Density 0.022%