INDEX

Explanations

prepositions followed by nouns

The neuron strongly activates on very common function words—especially the preposition “to” (and similar high-frequency words like “on,” “by,” “for,” “using,” 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

ڿ

-1.33

 extraordinaire

-1.31

Niveau

-1.30

 confé

-1.27

 abiye

-1.23

گوش

-1.20

 новом

-1.18

(≧

-1.17

 經典

-1.16

gratis

-1.15

POSITIVE LOGITS

 from

1.67

</b>

1.64

 into

1.49

 while

1.34

 first

1.22

</h2>

1.20

ЕР

1.16

SEGUIR

1.16

 through

1.14

 which

1.12

Activations Density 0.034%