INDEX

Explanations

javascript console logs

The neuron fires on tokens that look like programming code—e.g. function calls, variable names, JSON references, parentheses/brackets, and other code punctuation.

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

##

0.63

 représentation

0.61

Proposition

0.56

 मुस्

0.56

🦴

0.56

 गुण

0.54

 Carpathians

0.54

antikan

0.53

 গোলাপ

0.52

WORKDIR

0.52

POSITIVE LOGITS

 setInterval

1.45

 setTimeout

1.39

setTimeout

1.34

 XMLHttpRequest

1.18

var

1.11

getElementById

1.10

setInterval

1.08

 ReactDOM

1.03

 localStorage

1.02

clearInterval

1.02

Activations Density 0.009%