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

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

Neuron 4055 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: 0.2485. Min Range: -0.2485

Max Act: 0.2380. Min Act: -0.0001

Data Index: 1221367 (C4 (Web Text))

Max Activating Token Index: 84

Text #1

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2478. Min Act: -0.0001

Data Index: 1059271 (C4 (Web Text))

Max Activating Token Index: 959

Text #2

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2201. Min Act: -0.0001

Data Index: 1005030 (C4 (Web Text))

Max Activating Token Index: 288

Text #3

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2307. Min Act: -0.0001

Data Index: 561112 (C4 (Web Text))

Max Activating Token Index: 699

Text #4

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2073. Min Act: -0.0001

Data Index: 1062223 (C4 (Web Text))

Max Activating Token Index: 621

Text #5

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2485. Min Act: -0.0001

Data Index: 837635 (C4 (Web Text))

Max Activating Token Index: 345

Text #6

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2275. Min Act: -0.0001

Data Index: 627435 (C4 (Web Text))

Max Activating Token Index: 706

Text #7

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2259. Min Act: -0.0001

Data Index: 462055 (C4 (Web Text))

Max Activating Token Index: 997

Text #8

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2223. Min Act: -0.0000

Data Index: 268403 (C4 (Web Text))

Max Activating Token Index: 206

Text #9

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2195. Min Act: -0.0001

Data Index: 1449146 (Python Code)

Max Activating Token Index: 428

Text #10

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2209. Min Act: -0.0001

Data Index: 1065093 (C4 (Web Text))

Max Activating Token Index: 102

Text #11

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2025. Min Act: -0.0001

Data Index: 528487 (C4 (Web Text))

Max Activating Token Index: 235

Text #12

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2061. Min Act: -0.0001

Data Index: 450157 (C4 (Web Text))

Max Activating Token Index: 550

Text #13

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.1990. Min Act: -0.0001

Data Index: 1292005 (C4 (Web Text))

Max Activating Token Index: 941

Text #14

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2034. Min Act: -0.0001

Data Index: 924206 (C4 (Web Text))

Max Activating Token Index: 902

Text #15

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.1904. Min Act: -0.0001

Data Index: 1595275 (Python Code)

Max Activating Token Index: 235

Text #16

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2163. Min Act: -0.0001

Data Index: 717862 (C4 (Web Text))

Max Activating Token Index: 142

Text #17

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.1915. Min Act: -0.0000

Data Index: 139514 (C4 (Web Text))

Max Activating Token Index: 525

Text #18

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.2035. Min Act: -0.0000

Data Index: 1232956 (C4 (Web Text))

Max Activating Token Index: 344

Text #19

Max Range: 0.2485. Min Range: -0.2485

Max Act: 0.1840. Min Act: -0.0001

Data Index: 1630655 (Python Code)

Max Activating Token Index: 155