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

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

Neuron 2032 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: 1.5515. Min Range: -1.5515

Max Act: 1.5515. Min Act: -0.0001

Data Index: 478052 (C4 (Web Text))

Max Activating Token Index: 787

Text #1

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.4564. Min Act: -0.0001

Data Index: 1460385 (Python Code)

Max Activating Token Index: 400

Text #2

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.4024. Min Act: -0.0001

Data Index: 1070491 (C4 (Web Text))

Max Activating Token Index: 771

Text #3

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.3208. Min Act: -0.0001

Data Index: 1092720 (C4 (Web Text))

Max Activating Token Index: 161

Text #4

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.2857. Min Act: -0.0001

Data Index: 94075 (C4 (Web Text))

Max Activating Token Index: 774

Text #5

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.3389. Min Act: -0.0001

Data Index: 788406 (C4 (Web Text))

Max Activating Token Index: 220

Text #6

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.3121. Min Act: -0.0001

Data Index: 470311 (C4 (Web Text))

Max Activating Token Index: 790

Text #7

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.3483. Min Act: -0.0001

Data Index: 1433932 (Python Code)

Max Activating Token Index: 697

Text #8

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.3798. Min Act: -0.0001

Data Index: 403223 (C4 (Web Text))

Max Activating Token Index: 454

Text #9

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.2795. Min Act: -0.0001

Data Index: 803368 (C4 (Web Text))

Max Activating Token Index: 673

Text #10

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.3557. Min Act: -0.0001

Data Index: 253476 (C4 (Web Text))

Max Activating Token Index: 686

Text #11

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.3368. Min Act: -0.0001

Data Index: 899505 (C4 (Web Text))

Max Activating Token Index: 231

Text #12

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.2613. Min Act: -0.0001

Data Index: 1493970 (Python Code)

Max Activating Token Index: 20

Text #13

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.2199. Min Act: -0.0001

Data Index: 1468495 (Python Code)

Max Activating Token Index: 53

Text #14

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.2643. Min Act: -0.0001

Data Index: 355946 (C4 (Web Text))

Max Activating Token Index: 353

Text #15

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.2694. Min Act: -0.0001

Data Index: 1283199 (C4 (Web Text))

Max Activating Token Index: 590

Text #16

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.1670. Min Act: -0.0001

Data Index: 1223526 (C4 (Web Text))

Max Activating Token Index: 406

Text #17

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.2211. Min Act: -0.0001

Data Index: 287726 (C4 (Web Text))

Max Activating Token Index: 85

Text #18

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.1852. Min Act: -0.0001

Data Index: 1362897 (Python Code)

Max Activating Token Index: 4

Text #19

Max Range: 1.5515. Min Range: -1.5515

Max Act: 1.1777. Min Act: -0.0001

Data Index: 703227 (C4 (Web Text))

Max Activating Token Index: 586