INDEX

Explanations

tie, trouser, trout, tree

The neuron fires strongly on uncommon or specialized content words—especially proper names, brand/library names, and technical or rare nouns.

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

pecabe

-3.55

𓍊

-2.81

很

-2.45

ágenes

-2.36

 лучше

-2.33

 avec

-2.28

 Hinblick

-2.28

を作り

-2.27

 ausein

-2.27

-2.25

POSITIVE LOGITS

"},

3.11

2.86

2.72

).

2.55

 Preisen

2.45

</i>

2.45

荜

2.34

哢

2.31

ྜ

2.28

 несмотря

2.22

Activations Density 0.001%