INDEX

Explanations

scientific and programming expressions

The neuron is strongly activated by programming identifiers (e.g. variable, field, or class names) and their assignment syntax.

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

などと

-1.72

forgettable

-1.52

 знают

-1.41

ruly

-1.35

ご馳走

-1.34

 recopila

-1.32

stoppable

-1.27

᾿

-1.27

あんな

-1.26

 אבל

-1.25

POSITIVE LOGITS

 will

1.86

 which

1.59

on

1.38

Other

1.34

第五

1.30

GRAFÍA

1.30

 with

1.23

of

1.19

第四

1.19

has

1.16

Activations Density 0.049%