INDEX

Explanations

Senator and Representative titles

The neuron is identifying occurrences of congressional title abbreviations like “Rep.” and “Sen.”

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

</strong>

-2.13

rvats

-1.82

</em>

-1.78

boretum

-1.78

cleros

-1.76

 ヴィンテージ

-1.70

 skydd

-1.64

𝐆

-1.61

。

-1.59

 einzel

-1.59

POSITIVE LOGITS

foreignKey

1.83

1.74

镦

1.72

1.69

and

1.67

forKey

1.59

 such

1.58

'';

1.56

不堪

1.56

我了

1.55

Activations Density 0.020%