INDEX

Explanations

javascript variable declarations

The neuron strongly activates on the JavaScript “var” keyword, i.e. variable‐declaration tokens.

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

ícil

-1.52

皆さま

-1.50

匐

-1.46

什么样的

-1.46

牀

-1.42

さてさて

-1.40

 Saltar

-1.37

怎样的

-1.35

就是这样

-1.35

 some

-1.34

POSITIVE LOGITS

and

1.98

1.77

</h3>

1.65

 brancas

1.60

↵↵

1.43

↵

1.40

"./

1.33



1.31

 Prancis

1.31

'";

1.30

Activations Density 0.014%