INDEX

Explanations

networking and game development kit

The neuron fires on programming code tokens—especially API or class/function names and other technical identifiers.

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



-1.84

癭

-1.77

weile

-1.59

———-

-1.53

 オレンジ

-1.52

 !!!!!

-1.52

——–

-1.48

豨

-1.45

 Espagne

-1.44

睃

-1.44

POSITIVE LOGITS

who

2.31

 whose

1.72

 from

1.63

 through

1.54

 with

1.48

Even

1.43

que

1.42

 without

1.40

 helps

1.31

Activations Density 0.022%