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

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

Neuron 1036 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.0168. Min Range: -1.0168

Max Act: 1.0168. Min Act: -0.0001

Data Index: 738585 (C4 (Web Text))

Max Activating Token Index: 186

Text #1

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.8457. Min Act: -0.0001

Data Index: 738787 (C4 (Web Text))

Max Activating Token Index: 87

Text #2

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.8625. Min Act: -0.0001

Data Index: 515565 (C4 (Web Text))

Max Activating Token Index: 210

Text #3

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.8243. Min Act: -0.0001

Data Index: 906353 (C4 (Web Text))

Max Activating Token Index: 269

Text #4

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.5923. Min Act: -0.0001

Data Index: 700333 (C4 (Web Text))

Max Activating Token Index: 221

Text #5

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.6722. Min Act: -0.0001

Data Index: 986998 (C4 (Web Text))

Max Activating Token Index: 126

Text #6

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.6195. Min Act: -0.0001

Data Index: 736995 (C4 (Web Text))

Max Activating Token Index: 253

Text #7

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.5895. Min Act: -0.0001

Data Index: 366127 (C4 (Web Text))

Max Activating Token Index: 507

Text #8

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.5504. Min Act: -0.0001

Data Index: 601949 (C4 (Web Text))

Max Activating Token Index: 237

Text #9

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.4973. Min Act: -0.0001

Data Index: 1311466 (C4 (Web Text))

Max Activating Token Index: 846

Text #10

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.6204. Min Act: -0.0001

Data Index: 912493 (C4 (Web Text))

Max Activating Token Index: 670

Text #11

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.4996. Min Act: -0.0001

Data Index: 598561 (C4 (Web Text))

Max Activating Token Index: 327

Text #12

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.5764. Min Act: -0.0001

Data Index: 566228 (C4 (Web Text))

Max Activating Token Index: 290

Text #13

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.5206. Min Act: -0.0001

Data Index: 345732 (C4 (Web Text))

Max Activating Token Index: 558

Text #14

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.6880. Min Act: -0.0001

Data Index: 1013669 (C4 (Web Text))

Max Activating Token Index: 819

Text #15

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.4687. Min Act: -0.0001

Data Index: 1076668 (C4 (Web Text))

Max Activating Token Index: 335

Text #16

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.4827. Min Act: -0.0001

Data Index: 1324513 (C4 (Web Text))

Max Activating Token Index: 128

Text #17

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.4831. Min Act: -0.0001

Data Index: 1167652 (C4 (Web Text))

Max Activating Token Index: 1019

Text #18

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.4279. Min Act: -0.0001

Data Index: 168200 (C4 (Web Text))

Max Activating Token Index: 418

Text #19

Max Range: 1.0168. Min Range: -1.0168

Max Act: 0.5241. Min Act: -0.0001

Data Index: 724692 (C4 (Web Text))

Max Activating Token Index: 746