INDEX

Explanations

Characters engaging in relationships

The neuron strongly activates on words (and punctuation) that mark the start of new sentences (i.e. sentence-initial tokens).

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

if

-1.70

was

-1.52

當時

-1.27

had

-1.23

 today

-1.20

 whenever

-1.18

 that

-1.13

今でも

-1.11

at

-1.11

in

-1.10

POSITIVE LOGITS

 Empieza

1.27

 kullanılır

1.21

 yöntem

1.20

Conheça

1.20

 découvrir

1.19

 câmera

1.18

Outras

1.17

ynchronously

1.16

bagai

1.16

 Descubre

1.16

Activations Density 0.070%