INDEX

Explanations

newspapers, writers, columns

The neuron fires on explicit mentions of print‐news publications (e.g. “newspaper,” “newspapers,” “papers”).

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

</h2>

-2.28

que

-1.84

とその

-1.77

">

-1.73

aritions

-1.72

—

-1.70

ylase

-1.65

𓍊

-1.59

..."

-1.59

以及

-1.55

POSITIVE LOGITS

</i>

1.88

 newspapers

1.86

 newspaper

1.80

</u>

1.79

鱟

1.73

啣

1.70

哢

1.69

鰊

1.63

:";

1.63

羮

1.63

Activations Density 0.027%