INDEX

Explanations

checks for existence or state

This neuron detects negative-status or error-reporting phrases, such as “not found,” “is not set,” or “empty.”

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

of

-1.14

The

-1.08

and

-1.02

注意到

-1.02

そして

-1.00

without

-0.97

一趟

-0.97

 Japão

-0.94

 knowing

-0.94

But

-0.93

POSITIVE LOGITS

or

2.11

 или

1.49

1.38

 alebo

1.38

 یا

1.33

 поэтому

1.33

OrEmpty

1.32

或是

1.28

 please

1.24

หรือ

1.23

Activations Density 0.033%