INDEX

Explanations

precedes a summary

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>

2.13

↵↵

1.95

<start_of_image>

1.70

↵↵↵↵

1.46

↵↵↵↵↵

1.29

↵↵↵↵↵↵

1.26

↵↵↵

1.24

↵↵↵↵↵↵↵↵

1.13

 😉

1.10

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

1.06

POSITIVE LOGITS

</u>

1.61

</strong>

1.59

</h2>

1.57

</b>

1.50

</sub>

1.38

</h3>

1.29

</h5>

1.27

</h4>

1.27

】：

1.14

'})

1.11

Activations Density 0.062%