INDEX

Explanations

range of values

The neuron activates on words signaling numeric or allowable ranges (e.g. “range of,” “valid values,” “conditions”), i.e. it spots specifications of a range.

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

Κα

-0.89

 éte

-0.86

relenting

-0.84

ᚖ

-0.84

 maxed

-0.84

Khi

-0.83

Spinach

-0.83

Biografie

-0.83

seemed

-0.82

벤

-0.82

POSITIVE LOGITS

 possible

1.83

 acceptable

1.48

 applicability

1.43

 possibilities

1.42

 available

1.40

 values

1.30

 variation

1.29

 influence

1.27

 between

1.22

 activities

1.20

Activations Density 0.091%