INDEX

Explanations

in vitro

The neuron is detecting characters that indicate relationships between entities, particularly slashes, symbols, prepositions, and connecting words like "as" or "via."

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

and

-3.47

but

-2.63

 have

-2.44

 makes

-2.42

has

-2.36

 just

-2.36

 where

-2.30

 gives

-2.25

to

-2.20

by

-2.19

POSITIVE LOGITS

 rése

2.61

鲩

2.55

麃

2.36

 rigue

2.34

 dizaines

2.28

 auffi

2.25



2.25

𓆏

2.23

鱚

2.22

鹟

2.22

Activations Density 0.003%