INDEX

Explanations

monarchs and regnal numbers

The neuron strongly activates on capitalized personal names (i.e. individual people’s names) in the text.

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

sbereich

-1.84

 also

-1.64

especially

-1.60

either

-1.54

特に

-1.53

 Bersama

-1.49

 dergleichen

-1.49

.”

-1.47

 également

-1.45

</sup>

-1.45

POSITIVE LOGITS

of

2.78

知的

1.41

)=>{

1.41

 recinto

1.34

次代

1.34

ницей

1.34

 другие

1.31

solicited

1.30

ема

1.29

ENCIÓN

1.27

Activations Density 0.051%