Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

Dataset: 80% C4 (Web Text) and 20% Python Code

Neuron 1055 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 4.4338. Min Range: -4.4338

Max Act: 4.4338. Min Act: -0.0000

Data Index: 825334 (C4 (Web Text))

Max Activating Token Index: 148

Text #1

Max Range: 4.4338. Min Range: -4.4338

Max Act: 3.7767. Min Act: -0.0000

Data Index: 808430 (C4 (Web Text))

Max Activating Token Index: 565

Text #2

Max Range: 4.4338. Min Range: -4.4338

Max Act: 3.5400. Min Act: -0.0000

Data Index: 754273 (C4 (Web Text))

Max Activating Token Index: 918

Text #3

Max Range: 4.4338. Min Range: -4.4338

Max Act: 3.2398. Min Act: -0.0000

Data Index: 146021 (C4 (Web Text))

Max Activating Token Index: 906

Text #4

Max Range: 4.4338. Min Range: -4.4338

Max Act: 3.0151. Min Act: -0.0000

Data Index: 259554 (C4 (Web Text))

Max Activating Token Index: 274

Text #5

Max Range: 4.4338. Min Range: -4.4338

Max Act: 3.1765. Min Act: -0.0000

Data Index: 637894 (C4 (Web Text))

Max Activating Token Index: 962

Text #6

Max Range: 4.4338. Min Range: -4.4338

Max Act: 2.9767. Min Act: -0.0000

Data Index: 933759 (C4 (Web Text))

Max Activating Token Index: 865

Text #7

Max Range: 4.4338. Min Range: -4.4338

Max Act: 2.9149. Min Act: -0.0000

Data Index: 426218 (C4 (Web Text))

Max Activating Token Index: 943

Text #8

Max Range: 4.4338. Min Range: -4.4338

Max Act: 2.8549. Min Act: -0.0000

Data Index: 748232 (C4 (Web Text))

Max Activating Token Index: 885

Text #9

Max Range: 4.4338. Min Range: -4.4338

Max Act: 2.9650. Min Act: -0.0000

Data Index: 1256732 (C4 (Web Text))

Max Activating Token Index: 15

Text #10

Max Range: 4.4338. Min Range: -4.4338

Max Act: 3.1067. Min Act: -0.0000

Data Index: 190912 (C4 (Web Text))

Max Activating Token Index: 21

Text #11

Max Range: 4.4338. Min Range: -4.4338

Max Act: 2.8222. Min Act: -0.0000

Data Index: 1268528 (C4 (Web Text))

Max Activating Token Index: 46

Text #12

Max Range: 4.4338. Min Range: -4.4338

Max Act: 2.7879. Min Act: -0.0000

Data Index: 304767 (C4 (Web Text))

Max Activating Token Index: 166

Text #13

Max Range: 4.4338. Min Range: -4.4338

Max Act: 2.9198. Min Act: -0.0000

Data Index: 78382 (C4 (Web Text))

Max Activating Token Index: 60

Text #14

Max Range: 4.4338. Min Range: -4.4338

Max Act: 2.9568. Min Act: -0.0000

Data Index: 987751 (C4 (Web Text))

Max Activating Token Index: 948

Text #15

Max Range: 4.4338. Min Range: -4.4338

Max Act: 2.8550. Min Act: -0.0000

Data Index: 806955 (C4 (Web Text))

Max Activating Token Index: 759

Text #16

Max Range: 4.4338. Min Range: -4.4338

Max Act: 2.6348. Min Act: -0.0000

Data Index: 579676 (C4 (Web Text))

Max Activating Token Index: 905

Text #17

Max Range: 4.4338. Min Range: -4.4338

Max Act: 2.5640. Min Act: -0.0000

Data Index: 368151 (C4 (Web Text))

Max Activating Token Index: 463

Text #18

Max Range: 4.4338. Min Range: -4.4338

Max Act: 2.6395. Min Act: -0.0000

Data Index: 681697 (C4 (Web Text))

Max Activating Token Index: 578

Text #19

Max Range: 4.4338. Min Range: -4.4338

Max Act: 2.6385. Min Act: -0.0000

Data Index: 93085 (C4 (Web Text))

Max Activating Token Index: 356