INDEX

Explanations

programming initialization and operations

The neuron fires on imperative action words used as code identifiers—i.e. function or method names like “disconnect,” “disable,” “init,” “set,” and “draw.”

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

↵

-1.58

 jejich

-1.57

 ihrem

-1.50

 nachdem

-1.49

 conocemos

-1.44

等は

-1.42

良かったです

-1.41

 daraus

-1.40

 снова

-1.40

 again

-1.40

POSITIVE LOGITS

 them

1.98

 }));

1.73

();

1.66

1.63

 összefoglal

1.61

 őket

1.53

");

1.49

 segíts

1.48

鲼

1.46

Activations Density 0.021%