INDEX

Explanations

code structure and language

This neuron activates on programming‐language declaration and type‐specification keywords (e.g. “public,” “typename,” “in,” “out”).

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

 particolar

-1.76

 occasione

-1.68

ৄ

-1.66

 mesmas

-1.59

 virtù

-1.55

деров

-1.54

因為

-1.52

因为

-1.52

↵↵

-1.52

-1.51

POSITIVE LOGITS

and

2.67

”,

1.63

 from

1.56

();)

1.46

</h1>

1.39

lossians

1.38

ússia

1.38

入れて

1.38

us

1.37

ありますが

1.37

Activations Density 0.015%