INDEX

Explanations

a/an/the followed by noun classifier

The neuron strongly activates on the indefinite articles “a” and “an” (and their variants) when they introduce a noun phrase.

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

and

-3.41

on

-2.92

or

-2.91

 also

-2.66

to

-2.42

at

-2.30

not

-2.11

 when

-2.06

 During

-2.05

 with

-2.02

POSITIVE LOGITS

 decorativo

1.94

 erhi

1.91

 czarny

1.86

 œufs

1.78

凄く

1.77



1.77

素敵

1.76

 amazingly

1.73

 adhesivo

1.73

ほとん

1.72

Activations Density 0.052%