INDEX

Explanations

active actions

The neuron flags main action verbs that follow an intensifying adverb (e.g. “actively used,” “actively studied,” “actively developed,” “actively hiring”).

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

.......

-1.63

........

-1.62

 .....

-1.61

as

-1.52

 just

-1.43

.....

-1.41

 before

-1.38

 oddly

-1.38

-1.37

 when

-1.37

POSITIVE LOGITS

active

2.05

 active

1.63

actively

1.58

 актив

1.54

 ACTIVE

1.52

Active

1.49

 Active

1.41

 mauva

1.38

out

1.37

 rapi

1.30

Activations Density 0.034%