INDEX

Explanations

featuring artists and presenters

The neuron detects artist‐feature credit tokens (like “feat”, “ft”, “featuring” and their accompanying punctuation) in music track listings.

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

昻

-1.78

 jeśli

-1.75

flikt

-1.71

if

-1.62

 need

-1.52

⬯

-1.48

);

-1.48



-1.47

瓩

-1.45

-1.43

POSITIVE LOGITS

1.87

the

1.45

 véritable

1.40

 kollu

1.38

 Amongst

1.38

 esquecer

1.32

 humility

1.30

 nélk

1.29

 krem

1.27

徐々に

1.27

Activations Density 0.008%