INDEX

Explanations

/* and $

The neuron specializes in detecting standalone numeric tokens that represent the score (vote count) of a post.

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

for

-1.70

<bos>

-1.24

 four

-1.22

 three

-1.20

 with

-1.20

が

-1.19

を

-1.15

Sebelumnya

-1.14

 utterly

-1.13

six

-1.10

POSITIVE LOGITS



1.50

すっ

1.50

 !!!!

1.48

 nová

1.48

凄い

1.38

 VERY

1.37

色んな

1.37

exemple

1.34

</h4>

1.34

"'

1.34

Activations Density 0.060%