INDEX

Explanations

key principles or vulnerability occurs

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

 dropdown

0.96

 aguas

0.92

 unaff

0.91

 पदाधिकारियों

0.91

 carbono

0.89

浮

0.88

碳

0.88

mua

0.87

 trầm

0.86

 tín

0.85

POSITIVE LOGITS

 proofs

1.83

 Theorem

1.79

 theorem

1.76

 Theorems

1.74

 theorems

1.72

Theorem

1.70

 proof

1.65

 Proof

1.62

 Lemmas

1.62

Proofs

1.59

Activations Density 0.575%