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

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

Neuron 3613 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.7045. Min Range: -3.7045

Max Act: 3.7045. Min Act: -0.0001

Data Index: 1255868 (C4 (Web Text))

Max Activating Token Index: 118

Text #1

Max Range: 3.7045. Min Range: -3.7045

Max Act: 3.3179. Min Act: -0.0001

Data Index: 276276 (C4 (Web Text))

Max Activating Token Index: 220

Text #2

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.9300. Min Act: -0.0001

Data Index: 1030875 (C4 (Web Text))

Max Activating Token Index: 77

Text #3

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.7620. Min Act: -0.0001

Data Index: 506154 (C4 (Web Text))

Max Activating Token Index: 754

Text #4

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.9156. Min Act: -0.0001

Data Index: 835085 (C4 (Web Text))

Max Activating Token Index: 439

Text #5

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.6910. Min Act: -0.0001

Data Index: 1191718 (C4 (Web Text))

Max Activating Token Index: 432

Text #6

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.9367. Min Act: -0.0001

Data Index: 782300 (C4 (Web Text))

Max Activating Token Index: 359

Text #7

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.6262. Min Act: -0.0001

Data Index: 561776 (C4 (Web Text))

Max Activating Token Index: 752

Text #8

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.6403. Min Act: -0.0001

Data Index: 513300 (C4 (Web Text))

Max Activating Token Index: 850

Text #9

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.6819. Min Act: -0.0001

Data Index: 1059931 (C4 (Web Text))

Max Activating Token Index: 65

Text #10

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.6273. Min Act: -0.0001

Data Index: 82355 (C4 (Web Text))

Max Activating Token Index: 86

Text #11

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.5879. Min Act: -0.0001

Data Index: 187666 (C4 (Web Text))

Max Activating Token Index: 91

Text #12

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.4761. Min Act: -0.0000

Data Index: 1639614 (Python Code)

Max Activating Token Index: 896

Text #13

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.4212. Min Act: -0.0001

Data Index: 85845 (C4 (Web Text))

Max Activating Token Index: 215

Text #14

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.4329. Min Act: -0.0001

Data Index: 1300097 (C4 (Web Text))

Max Activating Token Index: 939

Text #15

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.2555. Min Act: -0.0001

Data Index: 1136324 (C4 (Web Text))

Max Activating Token Index: 853

Text #16

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.3460. Min Act: -0.0001

Data Index: 940226 (C4 (Web Text))

Max Activating Token Index: 327

Text #17

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.1393. Min Act: -0.0001

Data Index: 653040 (C4 (Web Text))

Max Activating Token Index: 967

Text #18

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.1769. Min Act: -0.0001

Data Index: 1164247 (C4 (Web Text))

Max Activating Token Index: 261

Text #19

Max Range: 3.7045. Min Range: -3.7045

Max Act: 2.3858. Min Act: -0.0001

Data Index: 782125 (C4 (Web Text))

Max Activating Token Index: 143