INDEX

Explanations

essential requirements and needs

The neuron flags technical, content-bearing words (especially verbs and key nouns) in academic or scientific prose.

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

 jarang

-0.90

 triomphe

-0.89

usters

-0.88

 pake

-0.86

 miniatura

-0.85

 différentes

-0.84

 AUTOR

-0.84

看了看

-0.84

 mépris

-0.83

 débit

-0.82

POSITIVE LOGITS

 required

1.13

 unim

1.03

 important

0.97

 needed

0.97

 properly

0.96

 freely

0.96

 during

0.96

 essential

0.94

 premise

0.94

 premises

0.90

Activations Density 0.089%