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

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

Neuron 4545 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.4868. Min Act: -0.0000

Data Index: 646104 (C4 (Web Text))

Max Activating Token Index: 520

Text #1

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.4872. Min Act: -0.0000

Data Index: 878385 (C4 (Web Text))

Max Activating Token Index: 1023

Text #2

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.4504. Min Act: -0.0000

Data Index: 934212 (C4 (Web Text))

Max Activating Token Index: 594

Text #3

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.4592. Min Act: 0.0000

Data Index: 1505529 (Python Code)

Max Activating Token Index: 119

Text #4

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3744. Min Act: -0.0000

Data Index: 278797 (C4 (Web Text))

Max Activating Token Index: 668

Text #5

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3788. Min Act: -0.0000

Data Index: 5493 (C4 (Web Text))

Max Activating Token Index: 249

Text #6

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3992. Min Act: -0.0000

Data Index: 1470198 (Python Code)

Max Activating Token Index: 773

Text #7

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.4345. Min Act: -0.0000

Data Index: 2998 (C4 (Web Text))

Max Activating Token Index: 604

Text #8

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3860. Min Act: -0.0000

Data Index: 823399 (C4 (Web Text))

Max Activating Token Index: 358

Text #9

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3853. Min Act: -0.0000

Data Index: 1327536 (C4 (Web Text))

Max Activating Token Index: 955

Text #10

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3658. Min Act: -0.0000

Data Index: 417651 (C4 (Web Text))

Max Activating Token Index: 246

Text #11

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3554. Min Act: -0.0000

Data Index: 56957 (C4 (Web Text))

Max Activating Token Index: 911

Text #12

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3411. Min Act: -0.0000

Data Index: 22111 (C4 (Web Text))

Max Activating Token Index: 399

Text #13

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3879. Min Act: -0.0000

Data Index: 803821 (C4 (Web Text))

Max Activating Token Index: 198

Text #14

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3155. Min Act: -0.0000

Data Index: 163862 (C4 (Web Text))

Max Activating Token Index: 514

Text #15

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3267. Min Act: -0.0000

Data Index: 1405392 (Python Code)

Max Activating Token Index: 457

Text #16

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3016. Min Act: -0.0000

Data Index: 1653038 (Python Code)

Max Activating Token Index: 471

Text #17

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3330. Min Act: -0.0000

Data Index: 1485183 (Python Code)

Max Activating Token Index: 799

Text #18

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3275. Min Act: -0.0000

Data Index: 664090 (C4 (Web Text))

Max Activating Token Index: 510

Text #19

Max Range: 0.4872. Min Range: -0.4872

Max Act: 0.3230. Min Act: -0.0000

Data Index: 602006 (C4 (Web Text))

Max Activating Token Index: 209