INDEX

Explanations

quantifier or progression

The neuron fires on small high-frequency function words across several languages (e.g. “al,” “e,” “no,” “ya,” “yang,” 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

and

-2.31

-2.23

 because

-2.02

but

-1.89

not

-1.76

 Since

-1.67

//

-1.63

an

-1.55

at

-1.53

 each

-1.52

POSITIVE LOGITS

聼

1.76

 aéreas

1.64

 láser

1.63

或者是

1.58

行

1.55

 colgante

1.53

segno

1.51

silhouette

1.51



1.50

瘁

1.48

Activations Density 0.004%