INDEX

Explanations

pre/post-verb/root patterns

The neuron strongly activates on tokens that include embedded numeric literals (digits and decimal points).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

㷫

0.58

DTDs

0.58

NAB

0.56

Stabilise

0.55

GPUs

0.55

EXAMPLES

0.54

BDF

0.54

ably

0.54

<unused973>

0.54

 Weddings

0.53

POSITIVE LOGITS

 septembre

0.73

 permet

0.72

 notific

0.68

 possibile

0.66

 anche

0.66

 effettu

0.65

 bisa

0.65

 diatas

0.64

р

0.64

 setelah

0.64

Activations Density 0.008%