INDEX

Explanations

ionization energy

The neuron is specialized to detect numeric tokens and quantitative values (e.g., numbers and measurements).

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

 Dostupné

-0.83

 Moussa

-0.79

 Judas

-0.78

 Adirond

-0.77

quets

-0.75

browse

-0.73

})();

-0.72

ifica

-0.72

чив

-0.71

"))

-0.69

POSITIVE LOGITS

 ionization

1.34

ionization

1.05

IE

0.92

IP

0.90

IE

0.88

removing

0.87

removal

0.85

Removal

0.84

 vertical

0.83

Removing

0.83

Activations Density 0.016%