INDEX

Explanations

if you were a flavor

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 وع

1.42

ческими

1.39

 меня

1.35

 conos

1.34

และความ

1.30

 интерес

1.29

 differenza

1.27

 относительно

1.26

 کچھ

1.26

 diferencias

1.24

POSITIVE LOGITS

ant

1.50

at

1.44

ion

1.38

atul

1.35

sof

1.34

 ofthe

1.28

fif

1.28

atani

1.25

chapters

1.21

 namesake

1.21

Activations Density 0.285%