INDEX

Explanations

labels and values

The neuron responds to multi‐digit numeric tokens (especially three-digit or larger numbers).

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

Ž

-1.30

Š

-1.22

#}

-1.17

tierten

-1.14

På

-1.12

</tfoot>

-1.10

 enligt

-1.08

+")

-1.08

Č

-1.07

+}\

-1.07

POSITIVE LOGITS

 trasforma

1.31

 esigen

1.30

があり

1.30

 facendo

1.27

TECN

1.19

って思

1.17

なども

1.14

 worin

1.11

 cercando

1.10

 poichè

1.09

Activations Density 0.036%