INDEX

Explanations

reflect and programming terms

The neuron fires on occurrences of the identifier “reflect” (e.g. import or package references to Go’s reflect library).

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

»,

-1.55

ischen

-1.41

‚

-1.41

xxv

-1.40

IER

-1.39

ious

-1.39

».

-1.37

reas

-1.36

preneurship

-1.35

ders

-1.33

POSITIVE LOGITS

 that

1.49

 this

1.39

of

1.30

 requires

1.26

it

1.20

 Якщо

1.19

who

1.19

 during

1.16

女友

1.14

now

1.13

Activations Density 0.023%