INDEX

Explanations

Friends of

This neuron detects occurrences of “Friends of …” group names (the phrase “Friends of” in organizational titles).

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

 делают

-0.82

zá

-0.79

Благо

-0.78

踴

-0.77

證明

-0.77

なくなる

-0.76

klis

-0.75

 efectiva

-0.75

 любовь

-0.74

makebox

-0.73

POSITIVE LOGITS

of

0.85

 like

0.84

 berat

0.79

dym

0.77

 فل

0.76

 валю

0.76

 garis

0.75

JMS

0.74

facet

0.73

藻

0.72

Activations Density 0.011%