INDEX

Explanations

complete and correct

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

 relentless

0.98

 seamless

0.95

 conveyor

0.92

少ない

0.91

 opravdu

0.89

 Lupus

0.89

 effortless

0.89

 très

0.87

 incredibly

0.87

 extremely

0.87

POSITIVE LOGITS

 okay

2.12

ok

2.09

OK

2.08

OK

1.90

 alright

1.86

Ok

1.84

Ok

1.82

 Okay

1.75

ok

1.70

okay

1.70

Activations Density 0.178%