INDEX

Explanations

ability to do things

The neuron strongly activates on content words that denote actions or shifts—verbs and related nouns like “attention,” “clears,” “identified,” “beyond,” “tell,” “operations,” “chance,” etc.—i.e. words indicating an action, change, or result.

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

 there

-1.78

֪

-1.42

 reciente

-1.38

 recientes

-1.36

 recientemente

-1.30

żdż

-1.26

鸢

-1.21

んでいる

-1.20

锄

-1.20

 terdapat

-1.20

POSITIVE LOGITS

1.50

1.31

1.29

1.18

entown

1.17

 still

1.17

 permite

1.16

Activations Density 0.851%