INDEX

Explanations

describing the process

This neuron primarily detects the pronoun “it,” especially when it appears as the subject at the start of a sentence.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

}()

0.30

ByteArray

0.29

その他の

0.26

}--\

0.25

}";

0.25

()}.

0.25

 znovu

0.24

selves

0.24

}^{*}\

0.24

}\

0.24

POSITIVE LOGITS

 inoltre

0.32

 consists

0.31

is

0.30

 fascinating

0.30

 همچنین

0.30

 ingenious

0.29

 astounding

0.29

 also

0.28

 بالإضافة

0.28

ological

0.27

Activations Density 0.422%