INDEX

Explanations

cylinder and words like it

This neuron detects occurrences of the root “cylind” (as in “cylinder,” “cylindrical,” etc.).

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

ICOLON

-0.81

투

-0.76

roch

-0.73

 Mante

-0.72

િ

-0.71

 viewModel

-0.70

 Азии

-0.70

fel

-0.70

adapters

-0.69

ghar

-0.69

POSITIVE LOGITS

Cyl

0.98

ྒ

0.88

ical

0.85

 runter

0.82

PON

0.81

 erfolgreich

0.80

ICAL

0.78

Hank

0.77

jScrollPane

0.75

éről

0.75

Activations Density 0.026%