INDEX

Explanations

world war or world within

This neuron responds to occurrences of the word “World,” typically as a capitalized term in titles or names.

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

 Moab

0.72

きた

0.72

 driveway

0.66

ರ

0.65

 simplex

0.64

PCP

0.59

 Forth

0.59

 potions

0.58

 Associates

0.58

 Memorandum

0.57

POSITIVE LOGITS

WIDE

1.05

वाइड

0.95

renowned

0.91

 renowned

0.85

widet

0.82

ளாவ

0.81

ワイド

0.80

各国

0.79

Wide

0.77

 renown

0.77

Activations Density 0.143%