INDEX

Explanations

code file headers

the word "file" in technical documentation and code comments, particularly after the "@file" tag.

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

‘

-1.80

 черная

-1.57

-1.51

-1.48

↵

-1.46

 comparable

-1.42

選べる

-1.38

-1.36

 plausible

-1.35

’:

-1.31

POSITIVE LOGITS

of

1.98

螽

1.66

ingabe

1.63

錞

1.57

亊

1.55

 bēr

1.49

 других

1.49

нг

1.48

siyon

1.47

BIÉN

1.45

Activations Density 0.003%