INDEX

Explanations

harmony and related concepts

The neuron detects occurrences of “harmon”—that is, words like harmony, harmonic, and harmonious.

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

Of

-2.89

 ของ

-2.59

 vecchia

-2.42

 nuova

-2.41

 bellissima

-2.39

裥

-2.31

 molta

-2.27

звичай

-2.27

 distanza

-2.27

 nuevos

-2.27

POSITIVE LOGITS

うーん

2.38

2.02

棪

2.00

1.97

↵

1.86

où

1.84

the

1.84

 děti

1.84

 sofá

1.82

of

1.81

Activations Density 0.016%