INDEX

Explanations

preposition or pronoun association

The neuron strongly activates on common English prepositions (e.g. “to,” “for,” “at,” “under,” “before”).

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

the

-1.73

to

-1.52

 from

-1.35

jango

-1.34

etero

-1.30

of

-1.29

pios

-1.29

 беже

-1.27

not

-1.24

-1.22

POSITIVE LOGITS

him

2.72

you

2.41

 them

2.16

us

1.82

me

1.81

leyin

1.47

 οποία

1.44

menü

1.40

 mình

1.37

 ekor

1.34

Activations Density 0.037%