INDEX

Explanations

either/or choices

The neuron fires on occurrences of the word “Either,” signaling the start of an either/or construction.

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

-2.50

-1.88

</h2>

-1.75

孩子们

-1.63

-1.59

莅

-1.59

其實

-1.55

 even

-1.54

");

-1.51

 حتی

-1.50

POSITIVE LOGITS

愎

1.90

 thing

1.88

 outspoken

1.71

she

1.70

 them

1.66

élène

1.61

์

1.57

🕗

1.57

もの

1.55

 immerse

1.53

Activations Density 0.014%