INDEX

Explanations

some sort ofthink about $ a monthenjoy thingswith argue youand dynamic

The neuron flags content‐bearing (non‐stop) words—i.e. the mid‐frequency nouns, verbs, and adjectives—while ignoring very common function words.

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

Trichloroethane

-1.27

tissement

-1.22

.}(

-1.10

着脸

-1.09

тельном

-1.05

って思

-1.02

 möjlighet

-1.02

 دارد

-1.01

*}\

-1.01

}}">

-1.00

POSITIVE LOGITS

 their

1.50

be

1.32

 anymore

1.27

1.22

1.15

 they

1.13

或是

1.11

airobi

1.10

1.09

Activations Density 0.013%