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

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

Neuron 3953 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: 1.5316. Min Range: -1.5316

Max Act: 1.5316. Min Act: -0.0001

Data Index: 1648328 (Python Code)

Max Activating Token Index: 1021

Text #1

Max Range: 1.5316. Min Range: -1.5316

Max Act: 1.2239. Min Act: -0.0001

Data Index: 1408657 (Python Code)

Max Activating Token Index: 323

Text #2

Max Range: 1.5316. Min Range: -1.5316

Max Act: 1.0514. Min Act: -0.0001

Data Index: 1437517 (Python Code)

Max Activating Token Index: 726

Text #3

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.9477. Min Act: -0.0001

Data Index: 1483144 (Python Code)

Max Activating Token Index: 205

Text #4

Max Range: 1.5316. Min Range: -1.5316

Max Act: 1.0359. Min Act: -0.0001

Data Index: 1604683 (Python Code)

Max Activating Token Index: 62

Text #5

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.9686. Min Act: -0.0001

Data Index: 1331115 (C4 (Web Text))

Max Activating Token Index: 657

Text #6

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.9304. Min Act: -0.0001

Data Index: 1441526 (Python Code)

Max Activating Token Index: 501

Text #7

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.8198. Min Act: -0.0001

Data Index: 631073 (C4 (Web Text))

Max Activating Token Index: 29

Text #8

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.9001. Min Act: -0.0001

Data Index: 1168528 (C4 (Web Text))

Max Activating Token Index: 946

Text #9

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.8118. Min Act: -0.0001

Data Index: 1155587 (C4 (Web Text))

Max Activating Token Index: 374

Text #10

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.8539. Min Act: -0.0001

Data Index: 1486764 (Python Code)

Max Activating Token Index: 158

Text #11

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.9699. Min Act: -0.0001

Data Index: 255600 (C4 (Web Text))

Max Activating Token Index: 155

Text #12

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.7456. Min Act: -0.0001

Data Index: 155113 (C4 (Web Text))

Max Activating Token Index: 653

Text #13

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.8380. Min Act: -0.0001

Data Index: 1370425 (Python Code)

Max Activating Token Index: 967

Text #14

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.8370. Min Act: -0.0001

Data Index: 998217 (C4 (Web Text))

Max Activating Token Index: 176

Text #15

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.8030. Min Act: -0.0001

Data Index: 1033073 (C4 (Web Text))

Max Activating Token Index: 857

Text #16

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.8284. Min Act: -0.0001

Data Index: 1626834 (Python Code)

Max Activating Token Index: 346

Text #17

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.7924. Min Act: -0.0001

Data Index: 1630637 (Python Code)

Max Activating Token Index: 496

Text #18

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.7997. Min Act: -0.0001

Data Index: 38333 (C4 (Web Text))

Max Activating Token Index: 837

Text #19

Max Range: 1.5316. Min Range: -1.5316

Max Act: 0.7997. Min Act: -0.0001

Data Index: 1537523 (Python Code)

Max Activating Token Index: 926