INDEX

Explanations

since its founding

This neuron strongly responds to commas and other punctuation or numeric tokens that signal clause breaks or statistical/quantitative segments.

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

is

-1.91

</h3>

-1.71

 same

-1.66

up

-1.45

chose

-1.43

-1.41

他没有

-1.38

所说

-1.36

因为它

-1.36

me

-1.33

POSITIVE LOGITS

 these

1.91

 this

1.79

our

1.62

 acoso

1.54

 இந்த

1.52

 Kepala

1.52

 poupée

1.50

 sebag

1.48

 memasak

1.45

обзор

1.43

Activations Density 0.063%