INDEX

Explanations

paths leading somewhere

path ahead

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

0.55

of

0.51

bf

0.50

 housewives

0.49

name

0.48

its

0.47

普段

0.47

رض

0.47

tet

0.46

NV

0.45

POSITIVE LOGITS

 path

1.48

 مسیر

1.40

 pathway

1.34

路径

1.29

 путь

1.29

 paths

1.28

 경로

1.27

 trajectory

1.27

 caminho

1.24

 பாதை

1.17

Activations Density 0.162%