INDEX

Explanations

nanny, babysitter, sitters, New York

The neuron strongly activates on occurrences of “baby” and “babysitter” tokens, so it’s detecting mentions of babysitting.

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

tedrale

-2.16

有多少

-1.77

狨

-1.69

to

-1.65

阗

-1.61

 traditionelle

-1.59

 geringe

-1.57

re

-1.57

 gesche

-1.55

珦

-1.55

POSITIVE LOGITS

⬮

1.85

 тук

1.81

?...

1.75

евра

1.69

𝙜

1.68

賅

1.67

煐

1.61

at

1.59

 sitter

1.58

íos

1.57

Activations Density 0.010%