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

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

Neuron 3460 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.1001. Min Range: -1.1001

Max Act: 1.1001. Min Act: -0.0000

Data Index: 1208859 (C4 (Web Text))

Max Activating Token Index: 565

Text #1

Max Range: 1.1001. Min Range: -1.1001

Max Act: 1.0155. Min Act: -0.0000

Data Index: 355859 (C4 (Web Text))

Max Activating Token Index: 795

Text #2

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.9532. Min Act: -0.0000

Data Index: 283248 (C4 (Web Text))

Max Activating Token Index: 39

Text #3

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.9123. Min Act: -0.0000

Data Index: 774173 (C4 (Web Text))

Max Activating Token Index: 731

Text #4

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.8650. Min Act: -0.0000

Data Index: 285214 (C4 (Web Text))

Max Activating Token Index: 593

Text #5

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.8371. Min Act: -0.0000

Data Index: 977424 (C4 (Web Text))

Max Activating Token Index: 82

Text #6

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.8314. Min Act: -0.0000

Data Index: 1640904 (Python Code)

Max Activating Token Index: 843

Text #7

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.7956. Min Act: -0.0000

Data Index: 780500 (C4 (Web Text))

Max Activating Token Index: 986

Text #8

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.8107. Min Act: -0.0000

Data Index: 1192619 (C4 (Web Text))

Max Activating Token Index: 430

Text #9

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.8016. Min Act: -0.0000

Data Index: 929412 (C4 (Web Text))

Max Activating Token Index: 360

Text #10

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.7480. Min Act: -0.0000

Data Index: 1000206 (C4 (Web Text))

Max Activating Token Index: 49

Text #11

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.7388. Min Act: -0.0000

Data Index: 1610558 (Python Code)

Max Activating Token Index: 957

Text #12

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.7077. Min Act: -0.0000

Data Index: 20386 (C4 (Web Text))

Max Activating Token Index: 251

Text #13

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.7091. Min Act: -0.0000

Data Index: 1635401 (Python Code)

Max Activating Token Index: 184

Text #14

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.7442. Min Act: -0.0000

Data Index: 1471192 (Python Code)

Max Activating Token Index: 683

Text #15

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.7681. Min Act: -0.0000

Data Index: 426538 (C4 (Web Text))

Max Activating Token Index: 832

Text #16

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.7193. Min Act: -0.0000

Data Index: 1413827 (Python Code)

Max Activating Token Index: 577

Text #17

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.7299. Min Act: -0.0000

Data Index: 812001 (C4 (Web Text))

Max Activating Token Index: 878

Text #18

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.7060. Min Act: -0.0000

Data Index: 287211 (C4 (Web Text))

Max Activating Token Index: 806

Text #19

Max Range: 1.1001. Min Range: -1.1001

Max Act: 0.7212. Min Act: -0.0000

Data Index: 728971 (C4 (Web Text))

Max Activating Token Index: 805