INDEX

Explanations

names and descriptions

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

 declarar

-0.81

 ſhould

-0.75

женская

-0.74

 vegetales

-0.73

 カメラ

-0.73

 library

-0.73

⼿

-0.73

 precip

-0.72

 возник

-0.72

דע

-0.72

POSITIVE LOGITS

COV

0.91

dropIfExists

0.85

">)</

0.84

 COVID

0.84

unknownFields

0.83

mlink

0.83

 τά

0.82

 ガー

0.81

 содержания

0.79

uliert

0.79

Activations Density 0.005%