INDEX

Explanations

Determine SpecificationsPractice Negative VisualizationReveal CharacterDon't Engage ImmediatelyDon't Get Bogged Down

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

<eos>

1.16

Mm

0.95

McE

0.94

mmas

0.92

:,

0.92

 Inference

0.92

Rum

0.92

 Survivors

0.91

 Surviv

0.91

iver

0.91

POSITIVE LOGITS

</h2>

1.60

</b>

1.55

</strong>

1.22

</h4>

1.19

</h3>

1.11

〟

1.07

﹞

1.06

</sub>

1.01

}$}

0.97

}`}

0.96

Activations Density 0.391%