INDEX

Explanations

date and time formats

The neuron fires on the colon separators in time or timestamp strings (e.g. the “:” between hours, minutes, and seconds).

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

in

-1.70

by

-1.55

of

-1.47

 with

-1.45

 that

-1.39

 ImageIcon

-1.34

 there

-1.21

 Desember

-1.20

 Januari

-1.20

 from

-1.19

POSITIVE LOGITS

 tumultuous

1.35

 luscious

1.30

 marvelous

1.28

 harrowing

1.25

coloriage

1.24

 commencer

1.23

Mvh

1.23

 delectable

1.21

 enraged

1.20

CurtirCurtir

1.20

Activations Density 0.006%