INDEX

Explanations

industrial and metals

The neuron detects mentions of weld-on mechanical parts and related strength claims.

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

 increíble

0.89

 प्रशंसा

0.88

 maravill

0.87

zięk

0.86

 oración

0.84

 улице

0.80

 अद्भुत

0.79

 siente

0.79

 swoim

0.78

 vicino

0.77

POSITIVE LOGITS

 mainly

1.00

 principally

0.95

 industrial

0.93

 chiefly

0.93

 largely

0.92

 industrialized

0.91

 metallurgy

0.86

主要是

0.84

 основных

0.82

 residues

0.81

Activations Density 0.147%