INDEX

Explanations

`|` and ` ` separating values

The neuron fires on isolated numeric tokens (e.g. table row IDs, counts, or other standalone numbers).

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

-1.59

zbęd

-1.59

 turisti

-1.55

powiedz

-1.52

趿

-1.52

 її

-1.52

Kalau

-1.49

絎

-1.47

zusammen

-1.47

 bluza

-1.47

POSITIVE LOGITS

2.05

1.59

</i>

1.55

</h3>

1.54

1.50

 have

1.50

 increased

1.48

 there

1.48

1.45

Activations Density 0.003%