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

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

Neuron 2070 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 8.3168. Min Range: -8.3168

Max Act: 8.1597. Min Act: -0.0000

Data Index: 886152 (C4 (Web Text))

Max Activating Token Index: 275

Text #1

Max Range: 8.3168. Min Range: -8.3168

Max Act: 8.3168. Min Act: -0.0000

Data Index: 1123023 (C4 (Web Text))

Max Activating Token Index: 730

Text #2

Max Range: 8.3168. Min Range: -8.3168

Max Act: 8.1981. Min Act: -0.0000

Data Index: 1112127 (C4 (Web Text))

Max Activating Token Index: 608

Text #3

Max Range: 8.3168. Min Range: -8.3168

Max Act: 8.0549. Min Act: -0.0000

Data Index: 928215 (C4 (Web Text))

Max Activating Token Index: 797

Text #4

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.7765. Min Act: -0.0000

Data Index: 376728 (C4 (Web Text))

Max Activating Token Index: 543

Text #5

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.3515. Min Act: -0.0000

Data Index: 1026558 (C4 (Web Text))

Max Activating Token Index: 137

Text #6

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.8230. Min Act: -0.0000

Data Index: 1130609 (C4 (Web Text))

Max Activating Token Index: 122

Text #7

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.7650. Min Act: -0.0000

Data Index: 995018 (C4 (Web Text))

Max Activating Token Index: 89

Text #8

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.4883. Min Act: -0.0000

Data Index: 848371 (C4 (Web Text))

Max Activating Token Index: 295

Text #9

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.8504. Min Act: -0.0000

Data Index: 817585 (C4 (Web Text))

Max Activating Token Index: 958

Text #10

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.6414. Min Act: -0.0000

Data Index: 346532 (C4 (Web Text))

Max Activating Token Index: 345

Text #11

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.4084. Min Act: -0.0000

Data Index: 490647 (C4 (Web Text))

Max Activating Token Index: 84

Text #12

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.1603. Min Act: -0.0000

Data Index: 14410 (C4 (Web Text))

Max Activating Token Index: 992

Text #13

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.0672. Min Act: -0.0000

Data Index: 288917 (C4 (Web Text))

Max Activating Token Index: 27

Text #14

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.0609. Min Act: -0.0000

Data Index: 102759 (C4 (Web Text))

Max Activating Token Index: 14

Text #15

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.2054. Min Act: -0.0000

Data Index: 541715 (C4 (Web Text))

Max Activating Token Index: 786

Text #16

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.1232. Min Act: -0.0000

Data Index: 1201490 (C4 (Web Text))

Max Activating Token Index: 308

Text #17

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.4221. Min Act: -0.0000

Data Index: 523977 (C4 (Web Text))

Max Activating Token Index: 217

Text #18

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.0268. Min Act: -0.0000

Data Index: 1269324 (C4 (Web Text))

Max Activating Token Index: 303

Text #19

Max Range: 8.3168. Min Range: -8.3168

Max Act: 7.1488. Min Act: -0.0000

Data Index: 903277 (C4 (Web Text))

Max Activating Token Index: 753