INDEX

Explanations

positive descriptors

The neuron fires on positive evaluative adjectives or adverbs—words used to praise or express approval.

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

-1.51

ymax

-1.38

 Leur

-1.32

}();

-1.31

autoIncrement

-1.30

ଃ

-1.28

：“……”

-1.26

 KNOW

-1.23

TryParse

-1.23

POSITIVE LOGITS

ücks

1.48

 různých

1.46

 maniere

1.38

жении

1.38

 jednod

1.37

 різних

1.36

豋

1.34

 fondi

1.34

 samoz

1.33

 dirinya

1.30

Activations Density 0.032%