INDEX

Explanations

first-line or outgoing

The neuron fires on adjectives that denote sequence or status (e.g. “first-line,” “second-generation,” “outgoing”).

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

 with

-1.88

違います

-1.38

 döne

-1.34

 just

-1.33

 таких

-1.31

 these

-1.29

 that

-1.26

larla

-1.25

poň

-1.23

футбо

-1.22

POSITIVE LOGITS

 remainder

1.63

 were

1.56

offers

1.43

有任何

1.42

considering

1.34

currently

1.34

but

1.34

рации

1.27

includes

1.25

 principi

1.25

Activations Density 0.106%