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

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

Neuron 3689 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.8794. Min Act: -0.0001

Data Index: 369115 (C4 (Web Text))

Max Activating Token Index: 658

Text #1

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.2811. Min Act: -0.0001

Data Index: 630295 (C4 (Web Text))

Max Activating Token Index: 647

Text #2

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.1333. Min Act: -0.0001

Data Index: 848175 (C4 (Web Text))

Max Activating Token Index: 279

Text #3

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.1631. Min Act: -0.0001

Data Index: 270376 (C4 (Web Text))

Max Activating Token Index: 111

Text #4

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.1733. Min Act: -0.0001

Data Index: 922349 (C4 (Web Text))

Max Activating Token Index: 592

Text #5

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.2952. Min Act: -0.0001

Data Index: 838782 (C4 (Web Text))

Max Activating Token Index: 263

Text #6

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.1984. Min Act: -0.0001

Data Index: 1314559 (C4 (Web Text))

Max Activating Token Index: 467

Text #7

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.0507. Min Act: -0.0001

Data Index: 199578 (C4 (Web Text))

Max Activating Token Index: 282

Text #8

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.1773. Min Act: -0.0001

Data Index: 115900 (C4 (Web Text))

Max Activating Token Index: 580

Text #9

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.1686. Min Act: -0.0001

Data Index: 1111404 (C4 (Web Text))

Max Activating Token Index: 920

Text #10

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.1379. Min Act: -0.0001

Data Index: 29287 (C4 (Web Text))

Max Activating Token Index: 584

Text #11

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.1608. Min Act: -0.0001

Data Index: 740192 (C4 (Web Text))

Max Activating Token Index: 292

Text #12

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.0696. Min Act: -0.0001

Data Index: 462411 (C4 (Web Text))

Max Activating Token Index: 298

Text #13

Max Range: 1.8794. Min Range: -1.8794

Max Act: 0.9225. Min Act: -0.0001

Data Index: 22387 (C4 (Web Text))

Max Activating Token Index: 826

Text #14

Max Range: 1.8794. Min Range: -1.8794

Max Act: 0.9963. Min Act: -0.0001

Data Index: 149076 (C4 (Web Text))

Max Activating Token Index: 1005

Text #15

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.1771. Min Act: -0.0001

Data Index: 965803 (C4 (Web Text))

Max Activating Token Index: 850

Text #16

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.1025. Min Act: -0.0001

Data Index: 1632412 (Python Code)

Max Activating Token Index: 840

Text #17

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.0948. Min Act: -0.0001

Data Index: 758539 (C4 (Web Text))

Max Activating Token Index: 140

Text #18

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.0536. Min Act: -0.0001

Data Index: 755837 (C4 (Web Text))

Max Activating Token Index: 328

Text #19

Max Range: 1.8794. Min Range: -1.8794

Max Act: 1.0022. Min Act: -0.0001

Data Index: 1371876 (Python Code)

Max Activating Token Index: 401