INDEX

Explanations

introducing categories or types

The neuron fires on enumeration and classification markers—numbers and related terms (e.g. “three,” “types,” “groups,” “categories,” “phases”) that introduce lists or classified items.

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

and

-2.38

 also

-1.49

not

-1.39

 could

-1.38

],

-1.34

had

-1.33

),

-1.20

 especially

-1.20

was

-1.17

 never

-1.13

POSITIVE LOGITS

1.80

 وهي

1.31

 thofe

1.30

 COMPOUNDS

1.29

 maig

1.29

 namely

1.27

 ideen

1.27

 abiye

1.27

分别是

1.27

 listes

1.26

Activations Density 0.067%