INDEX

Explanations

code or legal document citations

The neuron is detecting the legal citation marker “Slip Op” that appears in court‐opinion docket references.

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

を選択します

-1.42

 bicicletas

-1.39

and

-1.39

᯾

-1.26

者は

-1.23

 médic

-1.23

énus

-1.22

 compréhen

-1.21

 teclas

-1.21

 fhall

-1.17

POSITIVE LOGITS

to

1.35

<bos>

1.34

了几

1.29

但也

1.26

 mouseY

1.22

捗

1.20

着一个

1.15

⼿

1.13

見えます

1.12

ଗ

1.09

Activations Density 0.000%