INDEX

Explanations

phrases indicating need or strain

The neuron responds to formal, academic or technical wording—particularly common function and scholarly-content words (e.g. “to,” “a,” “the,” “development,” “enables,” “strain”) characteristic of scientific or formal prose.

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

it

-2.73

 this

-2.44

its

-2.42

 these

-2.08

 there

-1.80

 that

-1.71

امد

-1.41

これは

-1.24

 those

-1.23

Its

-1.23

POSITIVE LOGITS

 ซึ่ง

1.34

 OGSÅ

1.23

 lutego

1.21

僕

1.16

 totiž

1.16

 heilig

1.16

1.15

 یعنی

1.14

olkata

1.13

 Descripción

1.13

Activations Density 0.088%