INDEX

Explanations

is an understatement

The neuron fires on subjective evaluation words—adjectives or adverbs (like “apt,” “true,” “right,” “understatement,” “misnomer,” “accurate”) that qualify or judge a statement.

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

 brev

-1.28

 مثل

-1.25

 ancianos

-1.24

 cristianos

-1.22

 seperti

-1.22

</i>

-1.21

 jóvenes

-1.20

INA

-1.18

 enfermos

-1.18

</h5>

-1.17

POSITIVE LOGITS

 most

1.54

 probably

1.50

for

1.30

 probable

1.26

and

1.25

 بالنسبة

1.16

1.14

 that

1.13

 Probably

1.13

 hereafter

1.13

Activations Density 0.042%