INDEX

Explanations

superior companions, Greek origin, customer loyalty

This neuron fires strongly on the first key content word that signals the start of a new section or topic (e.g. “Chinese,” “good,” “talent,” “great,” “states,” “finish”).

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

 immunore

0.78

グレード

0.76

袂

0.76

 tetahi

0.75

送

0.74

グラム

0.73

少数

0.73

丳

0.73

étaires

0.73

桐

0.73

POSITIVE LOGITS

</h2>

1.03

</h3>

0.89

</h5>

0.88

</h4>

0.87

bø

0.84

</h1>

0.84

 Jahres

0.70

 Хар

0.69

ify

0.68

 Gere

0.68

Activations Density 0.000%