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

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

Neuron 5065 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.0792. Min Range: -1.0792

Max Act: 1.0149. Min Act: -0.0001

Data Index: 491101 (C4 (Web Text))

Max Activating Token Index: 591

Text #1

Max Range: 1.0792. Min Range: -1.0792

Max Act: 1.0792. Min Act: -0.0001

Data Index: 97143 (C4 (Web Text))

Max Activating Token Index: 135

Text #2

Max Range: 1.0792. Min Range: -1.0792

Max Act: 1.0535. Min Act: -0.0001

Data Index: 561105 (C4 (Web Text))

Max Activating Token Index: 772

Text #3

Max Range: 1.0792. Min Range: -1.0792

Max Act: 1.0157. Min Act: -0.0001

Data Index: 1252129 (C4 (Web Text))

Max Activating Token Index: 559

Text #4

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.9871. Min Act: -0.0001

Data Index: 943850 (C4 (Web Text))

Max Activating Token Index: 764

Text #5

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.9569. Min Act: -0.0001

Data Index: 377544 (C4 (Web Text))

Max Activating Token Index: 681

Text #6

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.9401. Min Act: -0.0001

Data Index: 293473 (C4 (Web Text))

Max Activating Token Index: 861

Text #7

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.9816. Min Act: -0.0001

Data Index: 628761 (C4 (Web Text))

Max Activating Token Index: 108

Text #8

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.8733. Min Act: -0.0001

Data Index: 300277 (C4 (Web Text))

Max Activating Token Index: 763

Text #9

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.9123. Min Act: -0.0001

Data Index: 815785 (C4 (Web Text))

Max Activating Token Index: 51

Text #10

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.9082. Min Act: -0.0001

Data Index: 1300582 (C4 (Web Text))

Max Activating Token Index: 272

Text #11

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.9014. Min Act: -0.0001

Data Index: 1470151 (Python Code)

Max Activating Token Index: 941

Text #12

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.9351. Min Act: -0.0001

Data Index: 435297 (C4 (Web Text))

Max Activating Token Index: 614

Text #13

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.9358. Min Act: -0.0001

Data Index: 172619 (C4 (Web Text))

Max Activating Token Index: 91

Text #14

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.8944. Min Act: -0.0001

Data Index: 415538 (C4 (Web Text))

Max Activating Token Index: 290

Text #15

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.8917. Min Act: -0.0001

Data Index: 718386 (C4 (Web Text))

Max Activating Token Index: 275

Text #16

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.9397. Min Act: -0.0001

Data Index: 80849 (C4 (Web Text))

Max Activating Token Index: 7

Text #17

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.9348. Min Act: -0.0001

Data Index: 881067 (C4 (Web Text))

Max Activating Token Index: 446

Text #18

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.9095. Min Act: -0.0001

Data Index: 464628 (C4 (Web Text))

Max Activating Token Index: 155

Text #19

Max Range: 1.0792. Min Range: -1.0792

Max Act: 0.8978. Min Act: -0.0001

Data Index: 1085869 (C4 (Web Text))

Max Activating Token Index: 38