INDEX

Explanations

court opinions

The neuron responds strongly to sequences of individual uppercase letters (e.g. “F I L E D,” “O P I N I O N,” or other header words spelled out letter by letter).

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

“

-1.70

</u>

-1.52

 another

-1.51

</h1>

-1.38

the

-1.34

Geo

-1.32

 Maple

-1.27

نور

-1.25

lija

-1.24

了什么

-1.24

POSITIVE LOGITS

 этом

1.36

 não

1.30

");

1.27

ными

1.27

 would

1.25

chenen

1.24

 aplica

1.20

不會

1.20

 цього

1.20

無し

1.20

Activations Density 0.061%