INDEX

Explanations

introducing context or definitions

The neuron is triggered by romanized non-English tokens—especially Japanese song or artist names and other transliterated titles.

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

if

-2.23

of

-1.91

if

-1.55

 إذا

-1.38

 suelen

-1.37

⍺

-1.36

 partner

-1.31

 including

-1.30

ви

-1.28

hauled

-1.27

POSITIVE LOGITS

 this

1.55

靥

1.37

 молока

1.29

 others

1.28

 Irán

1.27

ariats

1.27

 anest

1.25

ạch

1.25

法规

1.25

 monstru

1.23

Activations Density 0.023%