INDEX

Explanations

so...that clauses

The neuron strongly activates on single-word pronouns (e.g. “I,” “it,” “he,” “we,” “you”) used as subjects or discourse markers.

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

-1.88

 They

-1.41

-1.38

↵↵

-1.34

…

-1.33

 лента

-1.30

folios

-1.27

 Conversely

-1.25

 lakini

-1.23

POSITIVE LOGITS

⩖

1.55

 vasto

1.51

même

1.50

 baskılı

1.50

écri

1.48

 précie

1.48

 couvercle

1.45

becomes

1.45

 zelve

1.43

resulting

1.42

Activations Density 0.014%