INDEX

Explanations

scientific facts or principles

The neuron detects words and phrases that invoke scientific or medical authority—i.e. appeals to “scientific,” “medical,” “proven/established,” “data-backed,” “supreme knowledge,” etc.

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

ুখী

0.59

 χαρακ

0.59

ग्राम

0.56

 деревя

0.55

ूबर

0.55

 immagin

0.55

 locu

0.54

 апреле

0.54

 imag

0.53

 nomad

0.53

POSITIVE LOGITS

 scientific

2.44

 scientifically

2.36

科学

2.33

scientific

2.31

 wetensch

2.25

 wissenschaft

2.19

科學

2.19

Scientific

2.16

 과학

2.16

 Scientific

2.14

Activations Density 0.657%