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

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

Neuron 3053 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.7031. Min Act: -0.0000

Data Index: 614512 (C4 (Web Text))

Max Activating Token Index: 625

Text #1

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.5973. Min Act: -0.0000

Data Index: 1117934 (C4 (Web Text))

Max Activating Token Index: 503

Text #2

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.6239. Min Act: -0.0000

Data Index: 299018 (C4 (Web Text))

Max Activating Token Index: 309

Text #3

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.5445. Min Act: -0.0000

Data Index: 878992 (C4 (Web Text))

Max Activating Token Index: 910

Text #4

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.5320. Min Act: -0.0000

Data Index: 68658 (C4 (Web Text))

Max Activating Token Index: 699

Text #5

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.4897. Min Act: -0.0000

Data Index: 363460 (C4 (Web Text))

Max Activating Token Index: 991

Text #6

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.4045. Min Act: -0.0000

Data Index: 483060 (C4 (Web Text))

Max Activating Token Index: 133

Text #7

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.4405. Min Act: -0.0000

Data Index: 1322888 (C4 (Web Text))

Max Activating Token Index: 557

Text #8

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.3722. Min Act: -0.0000

Data Index: 294686 (C4 (Web Text))

Max Activating Token Index: 397

Text #9

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.4123. Min Act: -0.0000

Data Index: 260328 (C4 (Web Text))

Max Activating Token Index: 236

Text #10

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.2622. Min Act: -0.0000

Data Index: 228926 (C4 (Web Text))

Max Activating Token Index: 446

Text #11

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.2777. Min Act: -0.0000

Data Index: 951941 (C4 (Web Text))

Max Activating Token Index: 996

Text #12

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.2752. Min Act: -0.0000

Data Index: 1358576 (C4 (Web Text))

Max Activating Token Index: 390

Text #13

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.1593. Min Act: -0.0000

Data Index: 472672 (C4 (Web Text))

Max Activating Token Index: 805

Text #14

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.2198. Min Act: 0.0000

Data Index: 473442 (C4 (Web Text))

Max Activating Token Index: 67

Text #15

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.2201. Min Act: -0.0000

Data Index: 890693 (C4 (Web Text))

Max Activating Token Index: 368

Text #16

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.1807. Min Act: 0.0000

Data Index: 549092 (C4 (Web Text))

Max Activating Token Index: 542

Text #17

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.1848. Min Act: -0.0000

Data Index: 1551090 (Python Code)

Max Activating Token Index: 664

Text #18

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.1844. Min Act: -0.0000

Data Index: 738164 (C4 (Web Text))

Max Activating Token Index: 943

Text #19

Max Range: 0.7031. Min Range: -0.7031

Max Act: 0.1821. Min Act: -0.0000

Data Index: 959323 (C4 (Web Text))

Max Activating Token Index: 866