INDEX

Explanations

Christianity and other languages

This neuron detects numeric metadata (e.g. vote counts or karma scores) in Hacker News comment threads.

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

 शुर

-1.55

NOSIS

-1.46

 Кубка

-1.45

𝗼

-1.43

 pada

-1.41

เกง

-1.40

 lebih

-1.40

lllllll

-1.36

chea

-1.36

%">

-1.35

POSITIVE LOGITS

1.75

 cristianismo

1.59

 дѣ

1.52

 discípulos

1.31

 会場

1.29

 cristiana

1.27

 evitando

1.19

 掃除

1.18

就好了

1.16

 espirituales

1.16

Activations Density 0.064%