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

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

Neuron 1371 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.9557. Min Act: -0.0001

Data Index: 617293 (C4 (Web Text))

Max Activating Token Index: 823

Text #1

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.5433. Min Act: -0.0001

Data Index: 750661 (C4 (Web Text))

Max Activating Token Index: 63

Text #2

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.7311. Min Act: -0.0001

Data Index: 751242 (C4 (Web Text))

Max Activating Token Index: 280

Text #3

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.5145. Min Act: -0.0001

Data Index: 749025 (C4 (Web Text))

Max Activating Token Index: 870

Text #4

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.4445. Min Act: -0.0001

Data Index: 717809 (C4 (Web Text))

Max Activating Token Index: 864

Text #5

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.4631. Min Act: -0.0001

Data Index: 1334117 (C4 (Web Text))

Max Activating Token Index: 924

Text #6

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.3168. Min Act: -0.0001

Data Index: 1251140 (C4 (Web Text))

Max Activating Token Index: 262

Text #7

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.2823. Min Act: -0.0001

Data Index: 97194 (C4 (Web Text))

Max Activating Token Index: 850

Text #8

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.1624. Min Act: -0.0001

Data Index: 924009 (C4 (Web Text))

Max Activating Token Index: 992

Text #9

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.2551. Min Act: -0.0001

Data Index: 1163274 (C4 (Web Text))

Max Activating Token Index: 839

Text #10

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.2450. Min Act: -0.0001

Data Index: 312147 (C4 (Web Text))

Max Activating Token Index: 503

Text #11

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.2293. Min Act: -0.0001

Data Index: 1030614 (C4 (Web Text))

Max Activating Token Index: 891

Text #12

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.2255. Min Act: -0.0001

Data Index: 1304723 (C4 (Web Text))

Max Activating Token Index: 158

Text #13

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.2334. Min Act: -0.0001

Data Index: 1472297 (Python Code)

Max Activating Token Index: 68

Text #14

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.3533. Min Act: -0.0001

Data Index: 1153661 (C4 (Web Text))

Max Activating Token Index: 25

Text #15

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.3376. Min Act: -0.0001

Data Index: 839753 (C4 (Web Text))

Max Activating Token Index: 104

Text #16

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.1302. Min Act: -0.0001

Data Index: 165624 (C4 (Web Text))

Max Activating Token Index: 1002

Text #17

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.0688. Min Act: -0.0001

Data Index: 240267 (C4 (Web Text))

Max Activating Token Index: 654

Text #18

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.0446. Min Act: -0.0001

Data Index: 712651 (C4 (Web Text))

Max Activating Token Index: 415

Text #19

Max Range: 2.9557. Min Range: -2.9557

Max Act: 2.0656. Min Act: -0.0001

Data Index: 1242325 (C4 (Web Text))

Max Activating Token Index: 860