Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 3805 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.3823. Min Act: -0.0001

Data Index: 153703 (C4 (Web Text))

Max Activating Token Index: 246

Text #1

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.2133. Min Act: -0.0001

Data Index: 792392 (C4 (Web Text))

Max Activating Token Index: 208

Text #2

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.1774. Min Act: -0.0001

Data Index: 392204 (C4 (Web Text))

Max Activating Token Index: 117

Text #3

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.1612. Min Act: -0.0001

Data Index: 296382 (C4 (Web Text))

Max Activating Token Index: 314

Text #4

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.1612. Min Act: -0.0001

Data Index: 571613 (C4 (Web Text))

Max Activating Token Index: 659

Text #5

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.1263. Min Act: -0.0001

Data Index: 1341890 (C4 (Web Text))

Max Activating Token Index: 604

Text #6

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.0605. Min Act: -0.0001

Data Index: 258362 (C4 (Web Text))

Max Activating Token Index: 571

Text #7

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.1676. Min Act: -0.0001

Data Index: 268741 (C4 (Web Text))

Max Activating Token Index: 815

Text #8

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.1151. Min Act: -0.0001

Data Index: 682580 (C4 (Web Text))

Max Activating Token Index: 831

Text #9

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.0470. Min Act: -0.0001

Data Index: 1309642 (C4 (Web Text))

Max Activating Token Index: 189

Text #10

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.0664. Min Act: -0.0001

Data Index: 1121205 (C4 (Web Text))

Max Activating Token Index: 865

Text #11

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.0368. Min Act: -0.0001

Data Index: 1132048 (C4 (Web Text))

Max Activating Token Index: 179

Text #12

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.0331. Min Act: -0.0001

Data Index: 359052 (C4 (Web Text))

Max Activating Token Index: 63

Text #13

Max Range: 1.3823. Min Range: -1.3823

Max Act: 0.9834. Min Act: -0.0001

Data Index: 1257933 (C4 (Web Text))

Max Activating Token Index: 675

Text #14

Max Range: 1.3823. Min Range: -1.3823

Max Act: 0.9202. Min Act: -0.0001

Data Index: 1573016 (Python Code)

Max Activating Token Index: 488

Text #15

Max Range: 1.3823. Min Range: -1.3823

Max Act: 0.9490. Min Act: -0.0001

Data Index: 1480108 (Python Code)

Max Activating Token Index: 779

Text #16

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.0125. Min Act: -0.0001

Data Index: 913311 (C4 (Web Text))

Max Activating Token Index: 463

Text #17

Max Range: 1.3823. Min Range: -1.3823

Max Act: 1.0033. Min Act: -0.0001

Data Index: 741235 (C4 (Web Text))

Max Activating Token Index: 93

Text #18

Max Range: 1.3823. Min Range: -1.3823

Max Act: 0.9637. Min Act: -0.0001

Data Index: 1274941 (C4 (Web Text))

Max Activating Token Index: 381

Text #19

Max Range: 1.3823. Min Range: -1.3823

Max Act: 0.9910. Min Act: -0.0001

Data Index: 1038781 (C4 (Web Text))

Max Activating Token Index: 600