INDEX

Explanations

energy and nutrients

The main thing this neuron does is detect mentions of dieting, nutrition, and health‐related regimens.

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

 Immobilien

0.79

 Democr

0.77

 Unemployment

0.75

 контора

0.74

 Государ

0.73

 وقوع

0.73

 สนาม

0.73

 Gewerks

0.71

chandelier

0.70

 Seismic

0.70

POSITIVE LOGITS

 dietary

2.86

 nutritional

2.59

 diets

2.53

 nutrition

2.50

 foods

2.48

 diet

2.47

 Dietary

2.42

 nutritionist

2.33

营养

2.30

 calorie

2.27

Activations Density 1.202%