INDEX

Explanations

widespread occurrence

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 widely

-1.16

 widespread

-1.15

 popular

-1.09

 Widely

-1.08

popular

-1.01

 prevalent

-1.01

 prevailing

-0.93

InjectAttribute

-0.82

Popular

-0.82

httphttps

-0.81

POSITIVE LOGITS

 himſelf

0.59

 myſelf

0.59

 preſent

0.57

 themſelves

0.53

 houſe

0.52

 pleaſure

0.49

 ethnographic

0.48

 ſtate

0.48

 Nuevas

0.47

 purpoſe

0.47

Activations Density 0.098%