INDEX

Explanations

part of / role in

This neuron fires on constructions describing something as a significant part of something else (e.g. “a big/huge part of…”).

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

莛

-1.86

nisses

-1.77

This

-1.74

летия

-1.70

}{@

-1.67

</i>

-1.66

borsa

-1.66

GRAFÍA

-1.64

atement

-1.63

мест

-1.63

POSITIVE LOGITS

1.59

 letzte

1.59

 then

1.55

ドキドキ

1.55

aimeJ

1.54

<bos>

1.53

 dritte

1.52

푣

1.51

 esche

1.49

 nuov

1.48

Activations Density 0.014%