INDEX

Explanations

plays an important role

The neuron detects emphasis or evaluative markers—common intensifier words (e.g. important, well, such, critical) that qualify or stress a noun or role.

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

 sirop

-1.28

˃

-1.19

 testé

-1.18

 linkovi

-1.17

 politiker

-1.16

ᜀ

-1.15

 воспользоваться

-1.15

〜！

-1.14

itemBuilder

-1.14

℅

-1.13

POSITIVE LOGITS

1.37

1.29

1.27

了不少

1.24

 ทั้ง

1.23

了自己

1.20

1.17

 antwoord

1.14

 verhaal

1.13

Activations Density 0.214%