INDEX

Explanations

which was

The neuron fires on copular “was” constructions in relative or reporting clauses—especially the “which was …” pattern often introducing quoted statements, statistics, or descriptive information.

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

?)

-1.42

caping

-1.35

-1.30

 BECAUSE

-1.30

他

-1.27

だろう

-1.26

lllll

-1.23

 obligaciones

-1.21

 deceased

-1.20

 estadísticas

-1.20

POSITIVE LOGITS

 from

1.71

Politique

1.43

as

1.34

Initially

1.32

 this

1.32

so

1.31

 Initially

1.31

 still

1.29

 initially

1.27

 including

1.26

Activations Density 0.116%