INDEX

Explanations

down

The neuron fires on numeric tokens or number-like strings (e.g. measurements, statistics, dimensions).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

toBeTruthy

0.69

+}$,

0.64

^+$

0.63

(+

0.62

hø

0.62

Hoch

0.62

线上

0.61

hoog

0.61

+},

0.61

}_{+}\

0.60

POSITIVE LOGITS

 lower

1.63

 LOWER

1.61

ต่ำ

1.49

 پایین

1.47

 низ

1.45

LOW

1.45

低

1.45

 Lower

1.44

Lower

1.42

低い

1.42

Activations Density 0.598%