INDEX

Explanations

floating point or bit

The neuron detects technical terms that name numeric types or measurement units (e.g. “point,” “bit”).

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

 already

-1.34

 their

-1.23

 bereits

-1.20

and

-1.16

 very

-1.10

))->

-1.08

in

-1.07

on

-1.06

 investigated

-1.06

eloma

-1.05

POSITIVE LOGITS

<bos>

1.59

 contextLoads

1.55

もあって

1.45

学年

1.22

licans

1.20

 häufiger

1.16

もあり

1.13

 désolés

1.13

眉头

1.11

urop

1.11

Activations Density 0.001%