INDEX

Explanations

serialize, stringify, inspect

The neuron activates on method names used to serialize or convert data into string form (e.g. stringify, dumps, toString, toSource, toDateString).

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

圌

-1.46

埽

-1.43

}=\

-1.42

-1.39

 raiſ

-1.38

跺

-1.37

のための

-1.36

 také

-1.34

艄

-1.34

嚎

-1.33

POSITIVE LOGITS

of

1.92

珝

1.68

Surprisingly

1.66

 explí

1.63

 Anschließend

1.59

Recently

1.55

 lokalen

1.55

Since

1.52

Have

1.52

Just

1.51

Activations Density 0.017%