INDEX

Explanations

foreign names and words

This neuron spotlights subword tokens containing non-ASCII letters or diacritical marks (e.g., “ö,” “ǒ,” HTML accent codes).

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

</h3>

-1.60

-1.52

お値段

-1.45

.";

-1.43

 rapi

-1.41

アルミ

-1.38

!」

-1.35

žití

-1.34

 tava

-1.34

私たちは

-1.33

POSITIVE LOGITS

in

1.82

 again

1.80

oneer

1.63

 mantuvo

1.61

 jednym

1.59

ならではの

1.55

 vuelven

1.53

tampa

1.52

 podle

1.51

 理由

1.50

Activations Density 0.012%