Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

Dataset: The Pile

Neuron 4686 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l-pile')

Text #0

Max Range: 7.1682. Min Range: -7.1682

Max Act: 7.1244. Min Act: -0.0000

Data Index: 70805 (The Pile)

Max Activating Token Index: 144

Text #1

Max Range: 7.1682. Min Range: -7.1682

Max Act: 7.1682. Min Act: -0.0000

Data Index: 30917 (The Pile)

Max Activating Token Index: 241

Text #2

Max Range: 7.1682. Min Range: -7.1682

Max Act: 6.5318. Min Act: -0.0000

Data Index: 394352 (The Pile)

Max Activating Token Index: 300

Text #3

Max Range: 7.1682. Min Range: -7.1682

Max Act: 6.2934. Min Act: -0.0000

Data Index: 50153 (The Pile)

Max Activating Token Index: 161

Text #4

Max Range: 7.1682. Min Range: -7.1682

Max Act: 6.2934. Min Act: -0.0000

Data Index: 1662226 (The Pile)

Max Activating Token Index: 161

Text #5

Max Range: 7.1682. Min Range: -7.1682

Max Act: 6.8155. Min Act: -0.0000

Data Index: 992064 (The Pile)

Max Activating Token Index: 149

Text #6

Max Range: 7.1682. Min Range: -7.1682

Max Act: 4.0977. Min Act: -0.0000

Data Index: 1356060 (The Pile)

Max Activating Token Index: 293

Text #7

Max Range: 7.1682. Min Range: -7.1682

Max Act: 4.2201. Min Act: -0.0000

Data Index: 141873 (The Pile)

Max Activating Token Index: 359

Text #8

Max Range: 7.1682. Min Range: -7.1682

Max Act: 3.8091. Min Act: -0.0000

Data Index: 1104933 (The Pile)

Max Activating Token Index: 123

Text #9

Max Range: 7.1682. Min Range: -7.1682

Max Act: 3.4771. Min Act: -0.0000

Data Index: 1734719 (The Pile)

Max Activating Token Index: 684

Text #10

Max Range: 7.1682. Min Range: -7.1682

Max Act: 3.9111. Min Act: -0.0000

Data Index: 380639 (The Pile)

Max Activating Token Index: 279

Text #11

Max Range: 7.1682. Min Range: -7.1682

Max Act: 3.4297. Min Act: -0.0000

Data Index: 523776 (The Pile)

Max Activating Token Index: 242

Text #12

Max Range: 7.1682. Min Range: -7.1682

Max Act: 3.7738. Min Act: -0.0000

Data Index: 48044 (The Pile)

Max Activating Token Index: 1000

Text #13

Max Range: 7.1682. Min Range: -7.1682

Max Act: 3.7267. Min Act: -0.0000

Data Index: 1600984 (The Pile)

Max Activating Token Index: 955

Text #14

Max Range: 7.1682. Min Range: -7.1682

Max Act: 3.2982. Min Act: -0.0000

Data Index: 1064931 (The Pile)

Max Activating Token Index: 214

Text #15

Max Range: 7.1682. Min Range: -7.1682

Max Act: 3.2826. Min Act: -0.0000

Data Index: 1733289 (The Pile)

Max Activating Token Index: 494

Text #16

Max Range: 7.1682. Min Range: -7.1682

Max Act: 2.9732. Min Act: -0.0000

Data Index: 680517 (The Pile)

Max Activating Token Index: 924

Text #17

Max Range: 7.1682. Min Range: -7.1682

Max Act: 3.8281. Min Act: -0.0000

Data Index: 305220 (The Pile)

Max Activating Token Index: 982

Text #18

Max Range: 7.1682. Min Range: -7.1682

Max Act: 3.6556. Min Act: -0.0000

Data Index: 1767022 (The Pile)

Max Activating Token Index: 312

Text #19

Max Range: 7.1682. Min Range: -7.1682

Max Act: 3.5440. Min Act: -0.0000

Data Index: 768165 (The Pile)

Max Activating Token Index: 544