INDEX

Explanations

sunsets and sunrises

the neuron activates on words and numbers describing solar events and times of day—especially “sunrise,” “sunset,” “rises,” “sets,” and their associated timestamps.

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

without

-2.28

where

-1.98

 tanpa

-1.90

thought

-1.90

despite

-1.89

through

-1.89

ية

-1.86

ኧ

-1.79

 dommages

-1.77

during

-1.73

POSITIVE LOGITS

1.53

1.48

and

1.47

 sunsets

1.45

华为

1.45

1.44

To

1.43

 multifaceted

1.41

 gebouw

1.37

Activations Density 0.005%