INDEX

Explanations

several followed by plural nouns

The neuron activates on words expressing numeric quantities or vague counts (e.g. “several,” “three,” 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

了些

-1.81

or

-1.67

two

-1.65

了一个

-1.61

了两

-1.51

 folgte

-1.45

了一個

-1.40

了一下

-1.37

那些

-1.36

了一些

-1.34

POSITIVE LOGITS

of

2.95

 hundred

2.34

 dozen

2.30

 different

2.06

 times

1.90

 thousand

1.86

 types

1.71

 kinds

1.67

 iterations

1.56

ของ

1.50

Activations Density 0.055%