INDEX

Explanations

working

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

paravant

-0.77

 Anſ

-0.75

 purpoſe

-0.75

ыгана

-0.74

ArgsConstructor

-0.74

 automatiques

-0.73

 myſelf

-0.73

 Monfieur

-0.72

Дереккөздер

-0.71

 hés

-0.71

POSITIVE LOGITS

0.41

pyx

0.40

 operations

0.39

 known

0.39

cu

0.39

fas

0.37

リオン

0.36

::

0.35

JPanel

0.35

 legacy

0.34

Activations Density 0.098%