INDEX

Explanations

not categorized, have never, foundation reinforcement

The neuron fires on tokens appearing in numeric/statistical contexts (counts, ages, percentages, etc.).

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

そういえば

-1.38

 romero

-1.23

 レスト

-1.23

 ビキニ

-1.22

 rosario

-1.22

продам

-1.20

 italiens

-1.19

 esmeralda

-1.19

LEBR

-1.18

 alip

-1.17

POSITIVE LOGITS

asable

1.31

 основ

1.30

yny

1.21

ìn

1.20

跺

1.17

bbons

1.16

颯

1.15

owment

1.14

ą

1.13

</strong>

1.13

Activations Density 0.001%