INDEX

Explanations

end of sentence punctuation

end markers and structural elements

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

崇

0.37

垠

0.33

ッキリ

0.32

ᴬ

0.32

 కూడా

0.32

 стаўкі

0.31

 sectarian

0.31

性は

0.31

 Gál

0.31

 казіно

0.31

POSITIVE LOGITS

<eos>

1.11

0.78

<unused63>

0.66

</code>

0.60

<0x00>

0.59

៕

0.57

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

0.52

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

0.50

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

0.50

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

0.50

Activations Density 0.077%