INDEX

Explanations

version components and combinations

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

 secondly

-1.09

 Secondly

-1.00

Secondly

-0.95

 második

-0.88

wiście

-0.88

ishwa

-0.86

ipeline

-0.84

 imidlertid

-0.82

íč

-0.78

 вторая

-0.78

POSITIVE LOGITS

 third

1.14

 mixed

1.12

 combination

1.11

 combined

1.05

 fourth

0.98

 others

0.97

 combinations

0.96

mixed

0.96

 neutral

0.94

 their

0.91

Activations Density 0.099%