INDEX

Explanations

AI, OpenAI, NASA, Google, research sources

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

akos

0.91

佰

0.90

atypes

0.86

acs

0.86

ansh

0.85

acos

0.85

 necessitating

0.85

 substantially

0.84

 $\|\

0.83

ያስ

0.82

POSITIVE LOGITS

 टिश

0.89

++];

0.87

 tissue

0.86

ξει

0.85

 राठ

0.84

 retir

0.84

ίνουν

0.83

tissue

0.82

 метал

0.81

idina

0.81

Activations Density 0.030%