INDEX

Explanations

prepositional phrases and clauses

The neuron strongly activates on connective function words (prepositions and pronouns) that link clauses—tokens like “in,” “as,” “that,” “it,” etc.

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

its

-1.55

But

-1.45

to

-1.41

where

-1.41

te

-1.41

 leone

-1.40

our

-1.38

return

-1.38

even

-1.34

 lupo

-1.34

POSITIVE LOGITS

 visitas

1.34

тут

1.28

ksiä

1.27

 starb

1.27

 which

1.27

AND

1.27

recen

1.26

aktu

1.26

detek

1.25

 redak

1.24

Activations Density 0.121%