INDEX

Explanations

looping through lists

The neuron fires on occurrences of the word “each” (including in code identifiers like forEach).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 essendo

-1.30

 poiché

-1.28

=";

-1.27

 stehenden

-1.23

irled

-1.21

’).

-1.21

 when

-1.20

due

-1.17

 avendo

-1.17

屢

-1.16

POSITIVE LOGITS

it

1.54

not

1.27

1.20

 pourrait

1.17

 потом

1.16

each

1.13

 each

1.13

end

1.12

 Türkçe

1.08

おすすめ

1.07

Activations Density 0.079%