INDEX

Explanations

just, even, and including

The neuron detects common qualifier or emphasis words (adverbs like “just,” “even,” “including,” “not,” 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

<bos>

-1.46

 there

-1.41

of

-1.41

in

-1.34

 других

-1.26

чих

-1.16

 neler

-1.12

ëlle

-1.10

 jeweils

-1.09

 arşivlendi

-1.07

POSITIVE LOGITS

 piccoli

1.56

 that

1.53

 previously

1.49

its

1.31

fontStyle

1.30

 małe

1.30

 njego

1.20

 kleinen

1.18

 iniz

1.15

ℏ

1.14

Activations Density 0.050%