INDEX

Explanations

special characters in code

This neuron fires on standalone punctuation‐only tokens used as formatting or markup delimiters (e.g. lines of “+”, “=”, “.”, “>”, “====”, etc.).

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

鹀

-1.77

毟

-1.70

჻

-1.70

 相談

-1.69

 produkty

-1.63

 eixo

-1.62

 botão

-1.61

 firmy

-1.60

 listy

-1.58

 botón

-1.57

POSITIVE LOGITS

1.78

↵↵

1.59

are

1.58

OR

1.52

 will

1.51

NegativeButton

1.50

：

1.46

1.43

に関する

1.42

it

1.38

Activations Density 0.167%