INDEX

Explanations

spaces and earth

This neuron detects mentions of sentient species or races—especially the word “human” or “humans.”

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

</h4>

-1.43

 they

-1.41

</sup>

-1.34

↵↵↵↵

-1.27

 Mereka

-1.26

 zelo

-1.23

 espagnol

-1.22

 loro

-1.20

not

-1.20

↵↵↵

-1.20

POSITIVE LOGITS

 impulsar

1.45

 rescate

1.45

 anstatt

1.41

henswürdigkeiten

1.32

攫

1.30

NCIS

1.25

espoir

1.25

 techno

1.21

潟

1.21

 revés

1.20

Activations Density 0.014%