INDEX

Explanations

names beginning with "pri" or "ka", "sequ", "subset", "range"

The neuron fires strongly on tokens from non-English text (foreign-language words or abbreviations).

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

↵

-2.81

-2.34

);

-2.08

 ….

-1.99

 あります

-1.95

</b>

-1.93

重要的

-1.93

 genellikle

-1.90

 piensan

-1.88

 tacit

-1.84

POSITIVE LOGITS

蛱

2.91

鈇

2.86

𓆜

2.66

2.56

नलोड

2.56

abilities

2.53

䛗

2.52

神

2.31

ſſen

2.30

玙

2.30

Activations Density 0.001%