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

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

Neuron 1510 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.1687. Min Range: -5.1687

Max Act: 5.1687. Min Act: -0.0000

Data Index: 1400310 (Python Code)

Max Activating Token Index: 946

Text #1

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.9416. Min Act: -0.0000

Data Index: 305996 (C4 (Web Text))

Max Activating Token Index: 403

Text #2

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.1059. Min Act: -0.0000

Data Index: 1515584 (Python Code)

Max Activating Token Index: 188

Text #3

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.9988. Min Act: -0.0000

Data Index: 752766 (C4 (Web Text))

Max Activating Token Index: 947

Text #4

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.6051. Min Act: -0.0000

Data Index: 644722 (C4 (Web Text))

Max Activating Token Index: 393

Text #5

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.3070. Min Act: -0.0000

Data Index: 1628495 (Python Code)

Max Activating Token Index: 106

Text #6

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.8395. Min Act: -0.0000

Data Index: 462053 (C4 (Web Text))

Max Activating Token Index: 802

Text #7

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.5082. Min Act: 0.0000

Data Index: 43955 (C4 (Web Text))

Max Activating Token Index: 612

Text #8

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.2979. Min Act: -0.0000

Data Index: 808059 (C4 (Web Text))

Max Activating Token Index: 754

Text #9

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.6634. Min Act: -0.0000

Data Index: 468968 (C4 (Web Text))

Max Activating Token Index: 743

Text #10

Max Range: 5.1687. Min Range: -5.1687

Max Act: 3.9783. Min Act: -0.0000

Data Index: 1526965 (Python Code)

Max Activating Token Index: 72

Text #11

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.2302. Min Act: -0.0000

Data Index: 101361 (C4 (Web Text))

Max Activating Token Index: 1007

Text #12

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.3853. Min Act: -0.0000

Data Index: 590067 (C4 (Web Text))

Max Activating Token Index: 588

Text #13

Max Range: 5.1687. Min Range: -5.1687

Max Act: 3.9208. Min Act: -0.0000

Data Index: 551432 (C4 (Web Text))

Max Activating Token Index: 415

Text #14

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.0159. Min Act: -0.0000

Data Index: 1422577 (Python Code)

Max Activating Token Index: 186

Text #15

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.1905. Min Act: -0.0000

Data Index: 1426547 (Python Code)

Max Activating Token Index: 748

Text #16

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.1390. Min Act: -0.0000

Data Index: 1416681 (Python Code)

Max Activating Token Index: 780

Text #17

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.1275. Min Act: -0.0000

Data Index: 1475297 (Python Code)

Max Activating Token Index: 580

Text #18

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.1203. Min Act: -0.0000

Data Index: 12871 (C4 (Web Text))

Max Activating Token Index: 597

Text #19

Max Range: 5.1687. Min Range: -5.1687

Max Act: 4.0757. Min Act: -0.0000

Data Index: 1420336 (Python Code)

Max Activating Token Index: 629