INDEX

Explanations

The neuron detects snippets of programming code—identifiers, punctuation, brackets, operators, and boolean/null literals.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

classmethod

0.66

staticmethod

0.62

 tqdm

0.61

 kivy

0.59

uptools

0.59

 Flask

0.58

 Steinmann

0.57

 MNIST

0.56

 torchvision

0.55

shaded

0.55

POSITIVE LOGITS

 JavaScript

1.28

javascript

1.27

JavaScript

1.25

 eslint

1.22

 javascript

1.21

hasOwnProperty

1.16

 undefined

1.15

!==

1.13

JS

1.13

undefined

1.13

Activations Density 0.016%