INDEX

Explanations

javascript variable declarations

The neuron fires strongly on technical‐style nouns and identifiers (e.g. “Fields,” “object,” “ids,” “metric”) as well as compound product terms built around “lamin–” (e.g. “laminated,” “laminator,” “laminates”).

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

częściej

0.66

Якщо

0.66

 azimuthal

0.62

हायला

0.61

หาก

0.61

这也

0.61

ந்துள்ளது

0.60

如果我们

0.59

 jeśli

0.59

यदि

0.59

POSITIVE LOGITS

var

1.30

var

1.24

let

1.15

 ஒரு

0.92

const

0.91

 sebuah

0.91

 olan

0.87

 ஓர்

0.84

let

0.82

 ενός

0.81

Activations Density 0.007%