INDEX

Explanations

breaking rules and laws

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

 時計

-0.86

 diferenças

-0.85

 gospod

-0.78

век

-0.78

 isEqualToString

-0.77

ထ

-0.75

 piese

-0.72

 dégust

-0.72

 withdrawing

-0.71

 malheureux

-0.71

POSITIVE LOGITS

 breaking

4.25

breaking

3.72

 violating

3.58

Breaking

3.58

 violate

3.56

 Breaking

3.50

 break

3.47

 violation

3.31

 breaks

3.19

 violated

3.19

Activations Density 0.095%