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

Dataset: The Pile

Neuron 16 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.6115. Min Range: -5.6115

Max Act: 5.4320. Min Act: -0.0000

Data Index: 1397740 (The Pile)

Max Activating Token Index: 163

Text #1

Max Range: 5.6115. Min Range: -5.6115

Max Act: 5.4752. Min Act: -0.0000

Data Index: 446458 (The Pile)

Max Activating Token Index: 405

Text #2

Max Range: 5.6115. Min Range: -5.6115

Max Act: 5.6115. Min Act: -0.0000

Data Index: 67262 (The Pile)

Max Activating Token Index: 950

Text #3

Max Range: 5.6115. Min Range: -5.6115

Max Act: 5.2808. Min Act: -0.0000

Data Index: 1070646 (The Pile)

Max Activating Token Index: 778

Text #4

Max Range: 5.6115. Min Range: -5.6115

Max Act: 5.3972. Min Act: -0.0000

Data Index: 459266 (The Pile)

Max Activating Token Index: 306

Text #5

Max Range: 5.6115. Min Range: -5.6115

Max Act: 5.3110. Min Act: -0.0000

Data Index: 1950130 (The Pile)

Max Activating Token Index: 264

Text #6

Max Range: 5.6115. Min Range: -5.6115

Max Act: 5.4174. Min Act: -0.0000

Data Index: 681805 (The Pile)

Max Activating Token Index: 632

Text #7

Max Range: 5.6115. Min Range: -5.6115

Max Act: 4.9513. Min Act: -0.0000

Data Index: 44131 (The Pile)

Max Activating Token Index: 844

Text #8

Max Range: 5.6115. Min Range: -5.6115

Max Act: 4.5716. Min Act: -0.0000

Data Index: 507021 (The Pile)

Max Activating Token Index: 538

Text #9

Max Range: 5.6115. Min Range: -5.6115

Max Act: 4.3494. Min Act: -0.0000

Data Index: 425174 (The Pile)

Max Activating Token Index: 112

Text #10

Max Range: 5.6115. Min Range: -5.6115

Max Act: 4.6175. Min Act: -0.0000

Data Index: 1617719 (The Pile)

Max Activating Token Index: 269

Text #11

Max Range: 5.6115. Min Range: -5.6115

Max Act: 4.7953. Min Act: -0.0000

Data Index: 1992424 (The Pile)

Max Activating Token Index: 808

Text #12

Max Range: 5.6115. Min Range: -5.6115

Max Act: 4.3879. Min Act: -0.0000

Data Index: 1744874 (The Pile)

Max Activating Token Index: 354

Text #13

Max Range: 5.6115. Min Range: -5.6115

Max Act: 4.0929. Min Act: -0.0000

Data Index: 1897299 (The Pile)

Max Activating Token Index: 733

Text #14

Max Range: 5.6115. Min Range: -5.6115

Max Act: 4.0786. Min Act: -0.0000

Data Index: 1780863 (The Pile)

Max Activating Token Index: 699

Text #15

Max Range: 5.6115. Min Range: -5.6115

Max Act: 4.4418. Min Act: -0.0000

Data Index: 693477 (The Pile)

Max Activating Token Index: 894

Text #16

Max Range: 5.6115. Min Range: -5.6115

Max Act: 4.5941. Min Act: -0.0000

Data Index: 1233596 (The Pile)

Max Activating Token Index: 930

Text #17

Max Range: 5.6115. Min Range: -5.6115

Max Act: 4.2040. Min Act: -0.0000

Data Index: 182771 (The Pile)

Max Activating Token Index: 57

Text #18

Max Range: 5.6115. Min Range: -5.6115

Max Act: 4.2645. Min Act: -0.0000

Data Index: 402296 (The Pile)

Max Activating Token Index: 701

Text #19

Max Range: 5.6115. Min Range: -5.6115

Max Act: 4.0153. Min Act: -0.0000

Data Index: 1387165 (The Pile)

Max Activating Token Index: 217