INDEX

Explanations

code `from` or `import`

The neuron strongly fires on stray “Â” tokens (mis-encoded/mojibake characters), i.e. encoding artifacts.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 δεν

-2.89

vantaged

-2.89

medriver

-2.55

-2.52

 空港

-2.38

 был

-2.27

LoadScene

-2.27

mainAxisSize

-2.25

which

-2.22

 sebagainya

-2.22

POSITIVE LOGITS

</strong>

3.34

we

2.80

که

2.52

2.41

 présents

2.38

 internationalen

2.30

2.22

 They

2.20

 Former

2.19

说起

2.17

Activations Density 0.013%