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

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

Neuron 3840 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.4745. Min Act: 0.0000

Data Index: 1254534 (C4 (Web Text))

Max Activating Token Index: 564

Text #1

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.4240. Min Act: 0.0000

Data Index: 737564 (C4 (Web Text))

Max Activating Token Index: 187

Text #2

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.4327. Min Act: -0.0000

Data Index: 944785 (C4 (Web Text))

Max Activating Token Index: 687

Text #3

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.4157. Min Act: 0.0000

Data Index: 104224 (C4 (Web Text))

Max Activating Token Index: 972

Text #4

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3673. Min Act: -0.0000

Data Index: 138709 (C4 (Web Text))

Max Activating Token Index: 886

Text #5

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3778. Min Act: -0.0000

Data Index: 647349 (C4 (Web Text))

Max Activating Token Index: 630

Text #6

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3842. Min Act: 0.0003

Data Index: 1180144 (C4 (Web Text))

Max Activating Token Index: 114

Text #7

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3679. Min Act: -0.0000

Data Index: 1063462 (C4 (Web Text))

Max Activating Token Index: 87

Text #8

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3506. Min Act: -0.0000

Data Index: 415789 (C4 (Web Text))

Max Activating Token Index: 352

Text #9

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3433. Min Act: 0.0001

Data Index: 177361 (C4 (Web Text))

Max Activating Token Index: 365

Text #10

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3556. Min Act: -0.0000

Data Index: 154694 (C4 (Web Text))

Max Activating Token Index: 945

Text #11

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3455. Min Act: 0.0003

Data Index: 666189 (C4 (Web Text))

Max Activating Token Index: 89

Text #12

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3529. Min Act: -0.0000

Data Index: 444951 (C4 (Web Text))

Max Activating Token Index: 307

Text #13

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3188. Min Act: -0.0000

Data Index: 1119654 (C4 (Web Text))

Max Activating Token Index: 831

Text #14

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3408. Min Act: 0.0002

Data Index: 292219 (C4 (Web Text))

Max Activating Token Index: 980

Text #15

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3257. Min Act: 0.0000

Data Index: 151976 (C4 (Web Text))

Max Activating Token Index: 968

Text #16

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3263. Min Act: 0.0000

Data Index: 530442 (C4 (Web Text))

Max Activating Token Index: 532

Text #17

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3249. Min Act: -0.0000

Data Index: 20665 (C4 (Web Text))

Max Activating Token Index: 721

Text #18

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3238. Min Act: -0.0000

Data Index: 641912 (C4 (Web Text))

Max Activating Token Index: 949

Text #19

Max Range: 0.4745. Min Range: -0.4745

Max Act: 0.3590. Min Act: -0.0000

Data Index: 1270076 (C4 (Web Text))

Max Activating Token Index: 575