Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 2560 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.7419. Min Act: -0.0000

Data Index: 165808 (C4 (Web Text))

Max Activating Token Index: 420

Text #1

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.6892. Min Act: -0.0000

Data Index: 761812 (C4 (Web Text))

Max Activating Token Index: 241

Text #2

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.6612. Min Act: -0.0000

Data Index: 66765 (C4 (Web Text))

Max Activating Token Index: 534

Text #3

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.7506. Min Act: -0.0000

Data Index: 743062 (C4 (Web Text))

Max Activating Token Index: 791

Text #4

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.6246. Min Act: -0.0000

Data Index: 497591 (C4 (Web Text))

Max Activating Token Index: 180

Text #5

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.5513. Min Act: -0.0000

Data Index: 972835 (C4 (Web Text))

Max Activating Token Index: 809

Text #6

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.6078. Min Act: -0.0000

Data Index: 702170 (C4 (Web Text))

Max Activating Token Index: 101

Text #7

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.5915. Min Act: -0.0000

Data Index: 167469 (C4 (Web Text))

Max Activating Token Index: 143

Text #8

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.5662. Min Act: 0.0000

Data Index: 740781 (C4 (Web Text))

Max Activating Token Index: 363

Text #9

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.5959. Min Act: -0.0000

Data Index: 408673 (C4 (Web Text))

Max Activating Token Index: 657

Text #10

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.5444. Min Act: -0.0000

Data Index: 981960 (C4 (Web Text))

Max Activating Token Index: 399

Text #11

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.5089. Min Act: -0.0000

Data Index: 611327 (C4 (Web Text))

Max Activating Token Index: 310

Text #12

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.5584. Min Act: -0.0000

Data Index: 1433570 (Python Code)

Max Activating Token Index: 312

Text #13

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.5329. Min Act: -0.0000

Data Index: 294456 (C4 (Web Text))

Max Activating Token Index: 590

Text #14

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.4975. Min Act: 0.0000

Data Index: 367828 (C4 (Web Text))

Max Activating Token Index: 243

Text #15

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.5650. Min Act: -0.0000

Data Index: 334160 (C4 (Web Text))

Max Activating Token Index: 212

Text #16

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.4834. Min Act: -0.0000

Data Index: 621994 (C4 (Web Text))

Max Activating Token Index: 238

Text #17

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.5089. Min Act: -0.0000

Data Index: 725234 (C4 (Web Text))

Max Activating Token Index: 905

Text #18

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.4317. Min Act: -0.0000

Data Index: 99036 (C4 (Web Text))

Max Activating Token Index: 275

Text #19

Max Range: 0.7506. Min Range: -0.7506

Max Act: 0.5193. Min Act: -0.0000

Data Index: 730998 (C4 (Web Text))

Max Activating Token Index: 48