INDEX

Explanations

references to figures

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

 methods

-0.89

 ibang

-0.88

 manually

-0.86

ρέπει

-0.86

 ポケット

-0.85

 bagay

-0.83

Nép

-0.83



-0.82

κτη

-0.81

 Aufenthal

-0.81

POSITIVE LOGITS

 inset

1.11

 caption

1.05

subfigure

0.99

 shows

0.81

 showing

0.80

 panel

0.79

panels

0.77

してみると

0.76

 left

0.76

caption

0.76

Activations Density 0.008%