INDEX

Explanations

auxiliary verbs or modals

The neuron strongly activates on very common function words (e.g. pronouns, negations, high‐frequency particles).

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

ي

0.61

结尾

0.59

DeleteDialogOpen

0.59

к

0.55

ໄດ້

0.55

рус

0.55

cribing

0.54

чення

0.54

 লাগানো

0.54

POSITIVE LOGITS

 เพราะ

1.24

!,

1.20

 کیونکہ

1.20

 क्योंकि

1.19

 eftersom

1.19

 puisqu

1.16

 çünkü

1.16

 ибо

1.13

!”,

1.13

:).

1.12

Activations Density 0.313%