INDEX

Explanations

foreign language words

The neuron flags content-bearing “keyword” tokens—nouns, verbs, and technical terms—rather than function words.

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

0.79

_)

0.78

ت

0.77

вез

0.76

groomed

0.73

_-

0.73

dets

0.71

Ass

0.71

Cat

0.71

sau

0.71

POSITIVE LOGITS

ร่างกาย

0.94

 maksymal

0.88

是没有

0.88

 coba

0.87

 fotografías

0.85

 keltiradi

0.81

 mascota

0.81

 konsumen

0.81

唒

0.80

 activación

0.80

Activations Density 0.000%