INDEX

Explanations

ag followed by specific suffixes

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hler

-0.89

更に

-0.84

ół

-0.82

 further

-0.81

どうやら

-0.80

therosclerosis

-0.78

mahami

-0.77

 Satt

-0.75

FURTHER

-0.75

 furthermore

-0.75

POSITIVE LOGITS

Ag

1.45

ag

1.37

Ag

1.32

AG

1.13

Agn

1.01

这些

0.92

AGR

0.89

 Agenda

0.87

Agenda

0.85

Aging

0.85

Activations Density 0.056%