INDEX

Explanations

words and terms definitions

The neuron fires on single-word labels or category names—i.e. tokens that name classes, methods, terms, words, databases, etc.

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

是怎么回事

-1.67

and

-1.45

 bună

-1.23

是怎么

-1.20

 ervaren

-1.13

 verschill

-1.13

 overeen

-1.11

チョイス

-1.11

 pagină

-1.10

 diferite

-1.10

POSITIVE LOGITS

of

2.80

 that

1.62

 skog

1.42

1.37

 himmel

1.35

 alkoh

1.27

1.24

 musik

1.22

kyr

1.22

 män

1.20

Activations Density 0.090%