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

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

Neuron 3620 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: 1.4908. Min Range: -1.4908

Max Act: 1.4908. Min Act: -0.0001

Data Index: 1406346 (Python Code)

Max Activating Token Index: 140

Text #1

Max Range: 1.4908. Min Range: -1.4908

Max Act: 1.3981. Min Act: -0.0001

Data Index: 1554224 (Python Code)

Max Activating Token Index: 391

Text #2

Max Range: 1.4908. Min Range: -1.4908

Max Act: 1.2261. Min Act: -0.0001

Data Index: 320238 (C4 (Web Text))

Max Activating Token Index: 507

Text #3

Max Range: 1.4908. Min Range: -1.4908

Max Act: 1.3791. Min Act: -0.0001

Data Index: 1085165 (C4 (Web Text))

Max Activating Token Index: 656

Text #4

Max Range: 1.4908. Min Range: -1.4908

Max Act: 1.2046. Min Act: -0.0001

Data Index: 486792 (C4 (Web Text))

Max Activating Token Index: 588

Text #5

Max Range: 1.4908. Min Range: -1.4908

Max Act: 1.1324. Min Act: -0.0001

Data Index: 358201 (C4 (Web Text))

Max Activating Token Index: 442

Text #6

Max Range: 1.4908. Min Range: -1.4908

Max Act: 1.1436. Min Act: -0.0001

Data Index: 1590771 (Python Code)

Max Activating Token Index: 670

Text #7

Max Range: 1.4908. Min Range: -1.4908

Max Act: 1.0429. Min Act: -0.0001

Data Index: 125280 (C4 (Web Text))

Max Activating Token Index: 215

Text #8

Max Range: 1.4908. Min Range: -1.4908

Max Act: 0.9940. Min Act: -0.0001

Data Index: 952523 (C4 (Web Text))

Max Activating Token Index: 522

Text #9

Max Range: 1.4908. Min Range: -1.4908

Max Act: 1.0189. Min Act: -0.0001

Data Index: 1223349 (C4 (Web Text))

Max Activating Token Index: 505

Text #10

Max Range: 1.4908. Min Range: -1.4908

Max Act: 1.0482. Min Act: -0.0001

Data Index: 1248905 (C4 (Web Text))

Max Activating Token Index: 475

Text #11

Max Range: 1.4908. Min Range: -1.4908

Max Act: 0.9922. Min Act: -0.0001

Data Index: 262229 (C4 (Web Text))

Max Activating Token Index: 675

Text #12

Max Range: 1.4908. Min Range: -1.4908

Max Act: 1.0269. Min Act: -0.0001

Data Index: 1082827 (C4 (Web Text))

Max Activating Token Index: 391

Text #13

Max Range: 1.4908. Min Range: -1.4908

Max Act: 0.9363. Min Act: -0.0001

Data Index: 400040 (C4 (Web Text))

Max Activating Token Index: 855

Text #14

Max Range: 1.4908. Min Range: -1.4908

Max Act: 1.0498. Min Act: -0.0001

Data Index: 312586 (C4 (Web Text))

Max Activating Token Index: 257

Text #15

Max Range: 1.4908. Min Range: -1.4908

Max Act: 1.0699. Min Act: -0.0001

Data Index: 678403 (C4 (Web Text))

Max Activating Token Index: 172

Text #16

Max Range: 1.4908. Min Range: -1.4908

Max Act: 0.9464. Min Act: -0.0001

Data Index: 665972 (C4 (Web Text))

Max Activating Token Index: 144

Text #17

Max Range: 1.4908. Min Range: -1.4908

Max Act: 0.9657. Min Act: -0.0001

Data Index: 1154438 (C4 (Web Text))

Max Activating Token Index: 117

Text #18

Max Range: 1.4908. Min Range: -1.4908

Max Act: 0.9301. Min Act: -0.0001

Data Index: 887346 (C4 (Web Text))

Max Activating Token Index: 719

Text #19

Max Range: 1.4908. Min Range: -1.4908

Max Act: 0.9963. Min Act: -0.0001

Data Index: 1137771 (C4 (Web Text))

Max Activating Token Index: 770