Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 818 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.5102. Min Act: -0.0001

Data Index: 728635 (C4 (Web Text))

Max Activating Token Index: 493

Text #1

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.5306. Min Act: -0.0001

Data Index: 1135553 (C4 (Web Text))

Max Activating Token Index: 535

Text #2

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.4873. Min Act: -0.0001

Data Index: 1108660 (C4 (Web Text))

Max Activating Token Index: 961

Text #3

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.4771. Min Act: -0.0001

Data Index: 354679 (C4 (Web Text))

Max Activating Token Index: 415

Text #4

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.4575. Min Act: -0.0001

Data Index: 582093 (C4 (Web Text))

Max Activating Token Index: 961

Text #5

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.4546. Min Act: -0.0001

Data Index: 341003 (C4 (Web Text))

Max Activating Token Index: 36

Text #6

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.4706. Min Act: -0.0001

Data Index: 1074440 (C4 (Web Text))

Max Activating Token Index: 995

Text #7

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.4184. Min Act: -0.0000

Data Index: 1136774 (C4 (Web Text))

Max Activating Token Index: 160

Text #8

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.4493. Min Act: -0.0002

Data Index: 1646031 (Python Code)

Max Activating Token Index: 236

Text #9

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.3898. Min Act: -0.0002

Data Index: 266424 (C4 (Web Text))

Max Activating Token Index: 343

Text #10

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.3577. Min Act: -0.0001

Data Index: 775104 (C4 (Web Text))

Max Activating Token Index: 602

Text #11

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.4256. Min Act: -0.0002

Data Index: 1434631 (Python Code)

Max Activating Token Index: 572

Text #12

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.4110. Min Act: -0.0001

Data Index: 107427 (C4 (Web Text))

Max Activating Token Index: 974

Text #13

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.3664. Min Act: -0.0001

Data Index: 792632 (C4 (Web Text))

Max Activating Token Index: 471

Text #14

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.4220. Min Act: -0.0000

Data Index: 427167 (C4 (Web Text))

Max Activating Token Index: 958

Text #15

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.3813. Min Act: -0.0002

Data Index: 1425606 (Python Code)

Max Activating Token Index: 299

Text #16

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.3960. Min Act: -0.0001

Data Index: 1096996 (C4 (Web Text))

Max Activating Token Index: 204

Text #17

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.3976. Min Act: -0.0000

Data Index: 419243 (C4 (Web Text))

Max Activating Token Index: 335

Text #18

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.3673. Min Act: -0.0001

Data Index: 228916 (C4 (Web Text))

Max Activating Token Index: 875

Text #19

Max Range: 0.5306. Min Range: -0.5306

Max Act: 0.4182. Min Act: -0.0001

Data Index: 656039 (C4 (Web Text))

Max Activating Token Index: 573