INDEX

Explanations

list and dictionary structures

The neuron fires on the “Python:” tag and the code/question header line that follows it.

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

//$

-0.79

 Tirol

-0.75

'));

-0.73

 מלח

-0.73

ayanan

-0.72

)));

-0.71

 Freiw

-0.71

"";

-0.71

stornos

-0.70

%;

-0.70

POSITIVE LOGITS

 list

4.91

 lists

3.78

list

3.66

 List

3.44

List

3.17

列表

3.14

 lista

3.13

 список

3.08

 LIST

3.00

 리스트

2.77

Activations Density 0.167%