INDEX

Explanations

used for training or evaluation

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

‟

1.34

ʽ

1.33

intervals

1.29

''

1.28

èles

1.27

Bowling

1.27

arran

1.25

conv

1.25

‘‘

1.25

ˮ

1.25

POSITIVE LOGITS

 indirectly

1.59

 insign

1.57

 STILL

1.54

BURY

1.54

RENCE

1.52

 namesake

1.47

 라고

1.42

 MEAS

1.41

 ONLY

1.40

 ALSO

1.39

Activations Density 0.032%