INDEX

Explanations

acronyms and "edge"," ">"

The neuron responds to fully uppercase initialisms or acronyms (e.g. GDPR, IDF, IDEA, IE).

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

</h2>

-2.72

朦

-2.36

 каким

-2.34

腹部

-2.27

</sup>

-2.25

-2.22

 fino

-2.17

</td>

-2.17

ኧ

-2.16

 utilizar

-2.14

POSITIVE LOGITS

鱚

2.34

驀

2.30

慫

2.30

 aqueles

2.28

梣

2.17

 休閒

2.16

 alábbi

2.14

みを

2.14

 anún

2.14

猜测

2.11

Activations Density 0.001%