Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 2160 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.7860. Min Act: -0.0000

Data Index: 1186825 (C4 (Web Text))

Max Activating Token Index: 923

Text #1

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.7047. Min Act: -0.0000

Data Index: 695481 (C4 (Web Text))

Max Activating Token Index: 938

Text #2

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.5167. Min Act: -0.0000

Data Index: 472174 (C4 (Web Text))

Max Activating Token Index: 399

Text #3

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.1796. Min Act: -0.0000

Data Index: 1418404 (Python Code)

Max Activating Token Index: 281

Text #4

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.6746. Min Act: -0.0000

Data Index: 1078712 (C4 (Web Text))

Max Activating Token Index: 831

Text #5

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.4338. Min Act: -0.0000

Data Index: 1363476 (Python Code)

Max Activating Token Index: 627

Text #6

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.8393. Min Act: -0.0000

Data Index: 640594 (C4 (Web Text))

Max Activating Token Index: 301

Text #7

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.6178. Min Act: -0.0000

Data Index: 1571204 (Python Code)

Max Activating Token Index: 715

Text #8

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.1029. Min Act: -0.0000

Data Index: 253272 (C4 (Web Text))

Max Activating Token Index: 939

Text #9

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.3266. Min Act: -0.0000

Data Index: 256865 (C4 (Web Text))

Max Activating Token Index: 444

Text #10

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.0416. Min Act: -0.0001

Data Index: 1571109 (Python Code)

Max Activating Token Index: 651

Text #11

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.3759. Min Act: -0.0000

Data Index: 1173030 (C4 (Web Text))

Max Activating Token Index: 607

Text #12

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.3796. Min Act: -0.0000

Data Index: 491510 (C4 (Web Text))

Max Activating Token Index: 763

Text #13

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.3146. Min Act: -0.0000

Data Index: 1429358 (Python Code)

Max Activating Token Index: 457

Text #14

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.2227. Min Act: -0.0000

Data Index: 514140 (C4 (Web Text))

Max Activating Token Index: 832

Text #15

Max Range: 4.8393. Min Range: -4.8393

Max Act: 3.6798. Min Act: -0.0000

Data Index: 1653748 (Python Code)

Max Activating Token Index: 658

Text #16

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.0347. Min Act: -0.0000

Data Index: 1532432 (Python Code)

Max Activating Token Index: 723

Text #17

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.2078. Min Act: -0.0000

Data Index: 1564737 (Python Code)

Max Activating Token Index: 975

Text #18

Max Range: 4.8393. Min Range: -4.8393

Max Act: 4.0157. Min Act: -0.0000

Data Index: 1167978 (C4 (Web Text))

Max Activating Token Index: 197

Text #19

Max Range: 4.8393. Min Range: -4.8393

Max Act: 3.8296. Min Act: -0.0000

Data Index: 828630 (C4 (Web Text))

Max Activating Token Index: 650