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

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

Neuron 3691 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.9948. Min Act: -0.0000

Data Index: 1106976 (C4 (Web Text))

Max Activating Token Index: 221

Text #1

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.6997. Min Act: -0.0000

Data Index: 1364870 (Python Code)

Max Activating Token Index: 26

Text #2

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.6886. Min Act: -0.0000

Data Index: 1160651 (C4 (Web Text))

Max Activating Token Index: 80

Text #3

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.6591. Min Act: -0.0000

Data Index: 485449 (C4 (Web Text))

Max Activating Token Index: 911

Text #4

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.7954. Min Act: -0.0000

Data Index: 1376312 (Python Code)

Max Activating Token Index: 93

Text #5

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.4782. Min Act: -0.0000

Data Index: 1486445 (Python Code)

Max Activating Token Index: 311

Text #6

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.7389. Min Act: -0.0000

Data Index: 900226 (C4 (Web Text))

Max Activating Token Index: 681

Text #7

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.4693. Min Act: -0.0000

Data Index: 485548 (C4 (Web Text))

Max Activating Token Index: 273

Text #8

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.5867. Min Act: -0.0000

Data Index: 1559331 (Python Code)

Max Activating Token Index: 20

Text #9

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.4399. Min Act: -0.0000

Data Index: 387334 (C4 (Web Text))

Max Activating Token Index: 706

Text #10

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.6416. Min Act: -0.0000

Data Index: 1520475 (Python Code)

Max Activating Token Index: 64

Text #11

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.6229. Min Act: -0.0000

Data Index: 1571841 (Python Code)

Max Activating Token Index: 106

Text #12

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.6037. Min Act: -0.0000

Data Index: 1386400 (Python Code)

Max Activating Token Index: 557

Text #13

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.3113. Min Act: -0.0000

Data Index: 550014 (C4 (Web Text))

Max Activating Token Index: 855

Text #14

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.4205. Min Act: -0.0000

Data Index: 1010808 (C4 (Web Text))

Max Activating Token Index: 291

Text #15

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.3897. Min Act: -0.0000

Data Index: 414272 (C4 (Web Text))

Max Activating Token Index: 156

Text #16

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.3720. Min Act: -0.0000

Data Index: 387912 (C4 (Web Text))

Max Activating Token Index: 889

Text #17

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.2382. Min Act: -0.0000

Data Index: 963258 (C4 (Web Text))

Max Activating Token Index: 867

Text #18

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.2424. Min Act: -0.0000

Data Index: 452424 (C4 (Web Text))

Max Activating Token Index: 156

Text #19

Max Range: 2.9948. Min Range: -2.9948

Max Act: 2.3334. Min Act: -0.0000

Data Index: 371870 (C4 (Web Text))

Max Activating Token Index: 901