INDEX

Explanations

technical abbreviations and terms

The neuron strongly activates on numeric measurements and related technical punctuation (e.g. numbers, parentheses around numbers, slashes, arrows) rather than ordinary words.

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

=\""

-1.43

‟

-1.34

<tr>

-1.34

or

-1.30

EXPERIENCE

-1.23

𝜂

-1.19

‪

-1.18

erweise

-1.17

</h2>

-1.16

 Accesat

-1.16

POSITIVE LOGITS

・・・・・

1.45

in

1.37

 Notwendigkeit

1.35

人には

1.31

 Anforderungen

1.29

photographe

1.29

 表情

1.27

lossen

1.27

 Störungen

1.23

 trouxe

1.22

Activations Density 0.353%