INDEX

Explanations

candle holders and shades

The neuron detects mentions of lighting fixtures and their components (e.g. candles, chandeliers, lamps, lamp shades, lamp holders).

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

酹

-2.17

四处

-1.99

ari

-1.84

</td>

-1.77



-1.66

-1.65

ኧ

-1.64

мії

-1.59

 Entscheidung

-1.57

wiada

-1.57

POSITIVE LOGITS

</strong>

3.19

to

2.19

</em>

1.77

1.68

</h4>

1.56

baca

1.54

 stemming

1.53

secutions

1.52

for

1.51

৫

1.51

Activations Density 0.015%