INDEX

Explanations

heavy followed by descriptor

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

uvwxyz

2.43

тна

2.33

umu

2.30

su

2.24

swe

2.22

 fingert

2.21

৭০

2.20

tait

2.19

unan

2.19

ution

2.18

POSITIVE LOGITS

TOOLS

2.92

ment

2.82

нным

2.79

ногие

2.79

رحبا

2.75

mités

2.71

=?";

2.71

੍ਹ

2.69

нные

2.67

ें

2.66

Activations Density 0.016%