INDEX

Explanations

descriptive adjectives and their follow-ons

The neuron selectively fires on descriptive adjectives that convey size, scale, or dimensionality (e.g. “small,” “huge,” “compact,” “high-definition,” “3D,” “low-end,” 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

 diversos

-1.98

 distintos

-1.63

 diferentes

-1.42

いいのか

-1.38

いっ

-1.38

色んな

-1.35

 perceptible

-1.34

了很多

-1.34

 unterschied

-1.31

 crucial

-1.30

POSITIVE LOGITS

 ones

2.20

one

1.72

 like

1.39

 Popis

1.33

 with

1.32

 from

1.28

 them

1.26

 using

1.25

 fully

1.25

 utterly

1.24

Activations Density 0.073%