INDEX

Explanations

subjective perception

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

‘

1.40

’

1.34

leri

1.33

iscope

1.30

nas

1.25

‌های

1.24

selen

1.22

 tế

1.22

्ता

1.21

('.')[

1.21

POSITIVE LOGITS

conducting

1.90

 setContentView

1.90

ostructures

1.89

 convertView

1.86

форд

1.86

<unused1252>

1.85

াভাব

1.84

 intuitively

1.83

 hemicontinuous

1.83

确

1.80

Activations Density 0.193%