INDEX

Explanations

data format representation

The neuron activates on technical terms that denote data encodings or digital format specifications (e.g. base64, JSON, XML, punycode, 2D array dimensions, binary bit depths).

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

 Iowa

0.70

alagi

0.66

 umbrella

0.66

 immovable

0.66

 chamb

0.65

্নার

0.65

aft

0.64

oo

0.64

jur

0.64

 podmín

0.64

POSITIVE LOGITS

化的

1.08

Converter

1.07

Conversion

0.97

Convert

0.95

转换为

0.95

轉換

0.92

化

0.92

 representation

0.91

 convert

0.89

Representation

0.89

Activations Density 0.084%