INDEX

Explanations

one of the following

This neuron detects enumeration contexts, especially the phrase pattern “one of the following” in lists of options.

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

 kazak

-1.38

⎯

-1.37

 tenerife

-1.32

!—

-1.27



-1.25

 mallorca

-1.23

 zorro

-1.23

Linki

-1.22

ar

-1.21

......”

-1.20

POSITIVE LOGITS

or

1.65

 from

1.39

.*;

1.34

 recommandée

1.29

at

1.24

 burbujas

1.24

 współpracy

1.23

 pieles

1.21

 cápsulas

1.20

 bicicletas

1.20

Activations Density 0.088%