INDEX

Explanations

smokers or Dolby

This neuron strongly responds to occurrences of the subword “Dol” (as in “Dolby,” “Dolph,” “Dolwar,” “Dolinský,” etc.).

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

’

-3.22

 longterm

-2.67

 inconspicuous

-2.48

 andere

-2.45

 nonstop

-2.44

 unspeakable

-2.41

 dodgy

-2.41

龊

-2.33

 “[

-2.31

 Other

-2.30

POSITIVE LOGITS

2.55

2.44

ি

2.34

体

2.31

<h2>

2.30

䛗

2.28

ᨕ

2.25

臏

2.25

ון

2.19

of

2.16

Activations Density 0.005%