INDEX

Explanations

token `define` followed by other tokens

assignment operators in code, particularly the equals sign used for variable assignment.

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

of

-2.41

-1.52

 again

-1.34

to

-1.32

-1.29

-1.27

was

-1.27

in

-1.24

still

-1.24

-1.22

POSITIVE LOGITS

 %@",

1.45

 serwis

1.42

 COMERC

1.31

 acides

1.31

asties

1.31

了过去

1.30

 („

1.30

 cuestiones

1.27

 géant

1.27

 dificultades

1.25

Activations Density 0.010%