INDEX

Explanations

tam

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

Tam

-0.95

Tam

-0.90

tam

-0.90

 Normdatei

-0.86

ConstraintMaker

-0.85

 Manipulation

-0.82

 tampering

-0.81

 Manip

-0.77

TAM

-0.77

 manipulation

-0.76

POSITIVE LOGITS

my

0.54

ers

0.54

ed

0.54

er

0.53

}`).

0.52

ran

0.49

poco

0.49

med

0.48

 specchio

0.46

tors

0.46

Activations Density 0.047%