INDEX

Explanations

SQL statements or lists

The neuron responds strongly to tokens containing digits (i.e. numeric identifiers, version numbers, or other digit‐based strings).

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

中略

-1.54

 інших

-1.47

 vorigen

-1.27

 kecelakaan

-1.27

our

-1.24

Leírás

-1.24

詳しくは

-1.21

 других

-1.20

 especially

-1.20

ﾋ

-1.20

POSITIVE LOGITS

—

1.58

and

1.55

 sezon

1.37

☑

1.36

 tutur

1.35

 veren

1.33

€”

1.30

撄

1.30

 februari

1.27

的模样

1.27

Activations Density 0.128%