INDEX

Explanations

"Not" followed by doubt or qualification

The neuron activates on occurrences of the word “Not,” especially when used as a sentence‐ or phrase‐initial negation.

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

 poliuret

-1.30

 colina

-1.27

imize

-1.26

Ранее

-1.26

 撮り

-1.24

浓郁

-1.22

 maniere

-1.19

-```

-1.17

scaron

-1.17

其實

-1.16

POSITIVE LOGITS

to

2.22

 that

1.81

 even

1.71

 unless

1.65

 gonna

1.63

 until

1.59

 only

1.59

 sure

1.50

1.46

in

1.45

Activations Density 0.021%