INDEX

Explanations

asking for information

placeholders or fill-in-the-blanks

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

 catt

0.43

 molybdenum

0.41

 lattices

0.40

 capitals

0.40

 substrings

0.38

 casks

0.38

 denser

0.37

inl

0.37

隻

0.36

克

0.36

POSITIVE LOGITS

______

0.87

 ______

0.84

 ____________

0.80

 __________

0.80

 _____________

0.79

___________

0.79

____

0.78

 ________

0.78

_______

0.77

____________

0.76

Activations Density 0.020%