INDEX

Explanations

types of systems, devices, or vehicles

The neuron fires on technical multi-word category labels—especially phrases like “X recognition,” “X database,” “X system,” or “X device” (e.g. “handwriting recognition,” “image databases,” “neurological devices,” etc.).

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

it

-1.70

 only

-1.46

 also

-1.45

万美元

-1.36

-'

-1.20

嘅

-1.20

\\

-1.16

 from

-1.13

is

-1.13

ATION

-1.13

POSITIVE LOGITS

 that

1.50

 dacă

1.42

segna

1.36

 Pokud

1.32

 recientemente

1.27

 récemment

1.24

 kürzlich

1.24

 tumult

1.24

 někter

1.24

recently

1.23

Activations Density 0.387%