INDEX

Explanations

accomplished actions

The neuron fires strongly on core action verbs (e.g. insult, throw, get, stay, miss, break, hit, survive).

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

to

-2.27

 this

-1.70

or

-1.56

FOR

-1.29

QtCore

-1.28

 That

-1.25

ях

-1.20

≦)

-1.20

It

-1.19

opoulos

-1.18

POSITIVE LOGITS

 chaleco

1.60

㎉

1.57

 capucha

1.47

zwi

1.38

 astonishing

1.38

Spécifications

1.38

 diputados

1.36

 wonderfully

1.35

 boletín

1.34

 wonderful

1.33

Activations Density 0.018%