INDEX

Explanations

overclocking and debugging

The neuron responds to long, specialized technical or scientific terminology (multi-syllable words with Latin/Greek roots).

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

 autént

0.55

powerful

0.55

 силы

0.53

 deseos

0.53

 poderos

0.52

 poderosa

0.52

achtige

0.50

つの

0.50

Powerful

0.50

ObjData

0.50

POSITIVE LOGITS

出来的

0.87

isation

0.85

ization

0.84

 شدن

0.79

 کردن

0.77

过程中

0.77

 xong

0.73

ing

0.69

isatie

0.69

ización

0.68

Activations Density 0.608%