INDEX

Explanations

XML and code declarations

The neuron strongly activates on the commas used to separate numeric (especially hex) entries in an array or data‐literal initializer.

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

-1.67

-1.46

 последние

-1.38

できるだけ

-1.35

 иногда

-1.28

<<<<<<<

-1.26

给我们

-1.26

‘’

-1.26

rester

-1.23

⁴

-1.20

POSITIVE LOGITS

</b>

2.02

Достоинства

1.48

 februar

1.45

</h1>

1.39

 aufgrund

1.38

alami

1.38

咲

1.36

kontak

1.36

ramientas

1.35

例句

1.32

Activations Density 0.003%