Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 48 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0875. Min Act: -0.0003

Data Index: 546219 (C4 (Web Text))

Max Activating Token Index: 220

Text #1

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0841. Min Act: -0.0003

Data Index: 1102752 (C4 (Web Text))

Max Activating Token Index: 1007

Text #2

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0858. Min Act: -0.0004

Data Index: 492292 (C4 (Web Text))

Max Activating Token Index: 448

Text #3

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0867. Min Act: -0.0003

Data Index: 883989 (C4 (Web Text))

Max Activating Token Index: 77

Text #4

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0837. Min Act: -0.0003

Data Index: 576194 (C4 (Web Text))

Max Activating Token Index: 787

Text #5

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0831. Min Act: -0.0003

Data Index: 445766 (C4 (Web Text))

Max Activating Token Index: 429

Text #6

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0818. Min Act: -0.0003

Data Index: 119978 (C4 (Web Text))

Max Activating Token Index: 625

Text #7

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0840. Min Act: -0.0003

Data Index: 543481 (C4 (Web Text))

Max Activating Token Index: 791

Text #8

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0827. Min Act: -0.0003

Data Index: 442236 (C4 (Web Text))

Max Activating Token Index: 270

Text #9

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0810. Min Act: -0.0004

Data Index: 602415 (C4 (Web Text))

Max Activating Token Index: 70

Text #10

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0820. Min Act: -0.0004

Data Index: 47078 (C4 (Web Text))

Max Activating Token Index: 338

Text #11

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0792. Min Act: -0.0003

Data Index: 607873 (C4 (Web Text))

Max Activating Token Index: 392

Text #12

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0819. Min Act: -0.0004

Data Index: 968164 (C4 (Web Text))

Max Activating Token Index: 381

Text #13

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0801. Min Act: -0.0003

Data Index: 26950 (C4 (Web Text))

Max Activating Token Index: 235

Text #14

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0786. Min Act: -0.0003

Data Index: 1165848 (C4 (Web Text))

Max Activating Token Index: 292

Text #15

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0796. Min Act: -0.0002

Data Index: 977390 (C4 (Web Text))

Max Activating Token Index: 498

Text #16

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0805. Min Act: -0.0004

Data Index: 1027928 (C4 (Web Text))

Max Activating Token Index: 695

Text #17

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0776. Min Act: -0.0003

Data Index: 660247 (C4 (Web Text))

Max Activating Token Index: 174

Text #18

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0771. Min Act: -0.0003

Data Index: 927524 (C4 (Web Text))

Max Activating Token Index: 625

Text #19

Max Range: 0.0875. Min Range: -0.0875

Max Act: 0.0802. Min Act: -0.0003

Data Index: 478370 (C4 (Web Text))

Max Activating Token Index: 853