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

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

Neuron 3339 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.7037. Min Act: -0.0000

Data Index: 613570 (C4 (Web Text))

Max Activating Token Index: 258

Text #1

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.5896. Min Act: -0.0000

Data Index: 176863 (C4 (Web Text))

Max Activating Token Index: 279

Text #2

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.5536. Min Act: -0.0000

Data Index: 327669 (C4 (Web Text))

Max Activating Token Index: 789

Text #3

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.4564. Min Act: -0.0000

Data Index: 465357 (C4 (Web Text))

Max Activating Token Index: 921

Text #4

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.3719. Min Act: -0.0000

Data Index: 762053 (C4 (Web Text))

Max Activating Token Index: 712

Text #5

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.5317. Min Act: -0.0000

Data Index: 1413059 (Python Code)

Max Activating Token Index: 614

Text #6

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.5156. Min Act: -0.0000

Data Index: 1368641 (Python Code)

Max Activating Token Index: 449

Text #7

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.4223. Min Act: -0.0000

Data Index: 1430659 (Python Code)

Max Activating Token Index: 642

Text #8

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.4039. Min Act: -0.0000

Data Index: 1534227 (Python Code)

Max Activating Token Index: 1008

Text #9

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.3075. Min Act: -0.0000

Data Index: 1609442 (Python Code)

Max Activating Token Index: 950

Text #10

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.3115. Min Act: -0.0000

Data Index: 1275645 (C4 (Web Text))

Max Activating Token Index: 320

Text #11

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.2854. Min Act: -0.0000

Data Index: 168918 (C4 (Web Text))

Max Activating Token Index: 678

Text #12

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.1537. Min Act: -0.0000

Data Index: 156918 (C4 (Web Text))

Max Activating Token Index: 126

Text #13

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.2074. Min Act: -0.0000

Data Index: 1509364 (Python Code)

Max Activating Token Index: 605

Text #14

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.1346. Min Act: -0.0000

Data Index: 427300 (C4 (Web Text))

Max Activating Token Index: 984

Text #15

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.2055. Min Act: -0.0000

Data Index: 147789 (C4 (Web Text))

Max Activating Token Index: 81

Text #16

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.2561. Min Act: -0.0000

Data Index: 287528 (C4 (Web Text))

Max Activating Token Index: 780

Text #17

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.2337. Min Act: 0.0000

Data Index: 357600 (C4 (Web Text))

Max Activating Token Index: 338

Text #18

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.1566. Min Act: 0.0000

Data Index: 395467 (C4 (Web Text))

Max Activating Token Index: 569

Text #19

Max Range: 1.7037. Min Range: -1.7037

Max Act: 1.1505. Min Act: -0.0000

Data Index: 1226523 (C4 (Web Text))

Max Activating Token Index: 76