INDEX

Explanations

correction

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

 correction

-1.08

corrected

-1.05

 corrected

-1.00

correction

-0.97

 Correction

-0.85

Correction

-0.84

 Corrected

-0.82

Corrected

-0.77

Correct

-0.71

 CORRE

-0.71

POSITIVE LOGITS

Efq

0.79

 mergeFrom

0.78

 raiſ

0.77

toMatchSnapshot

0.76

 Monfieur

0.74

 Majefty

0.73

 vuitton

0.73

contentLoaded

0.73

 perſon

0.73

 насељу

0.72

Activations Density 0.204%