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

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

Neuron 1642 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.3853. Min Act: -0.0001

Data Index: 448718 (C4 (Web Text))

Max Activating Token Index: 258

Text #1

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.3356. Min Act: -0.0001

Data Index: 1500656 (Python Code)

Max Activating Token Index: 780

Text #2

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.2765. Min Act: -0.0001

Data Index: 197412 (C4 (Web Text))

Max Activating Token Index: 869

Text #3

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.2187. Min Act: -0.0000

Data Index: 983522 (C4 (Web Text))

Max Activating Token Index: 64

Text #4

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.1582. Min Act: -0.0000

Data Index: 1217906 (C4 (Web Text))

Max Activating Token Index: 79

Text #5

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.2283. Min Act: -0.0001

Data Index: 364474 (C4 (Web Text))

Max Activating Token Index: 501

Text #6

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.2511. Min Act: -0.0000

Data Index: 785087 (C4 (Web Text))

Max Activating Token Index: 387

Text #7

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.2220. Min Act: -0.0001

Data Index: 3944 (C4 (Web Text))

Max Activating Token Index: 300

Text #8

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.0688. Min Act: -0.0001

Data Index: 476172 (C4 (Web Text))

Max Activating Token Index: 638

Text #9

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.2805. Min Act: -0.0001

Data Index: 1057941 (C4 (Web Text))

Max Activating Token Index: 657

Text #10

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.2507. Min Act: -0.0001

Data Index: 475955 (C4 (Web Text))

Max Activating Token Index: 213

Text #11

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.2053. Min Act: -0.0001

Data Index: 1386483 (Python Code)

Max Activating Token Index: 233

Text #12

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.1471. Min Act: -0.0000

Data Index: 1245326 (C4 (Web Text))

Max Activating Token Index: 80

Text #13

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.1528. Min Act: -0.0000

Data Index: 500132 (C4 (Web Text))

Max Activating Token Index: 312

Text #14

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.2089. Min Act: -0.0000

Data Index: 455438 (C4 (Web Text))

Max Activating Token Index: 416

Text #15

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.1583. Min Act: -0.0000

Data Index: 482527 (C4 (Web Text))

Max Activating Token Index: 391

Text #16

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.1316. Min Act: -0.0001

Data Index: 181652 (C4 (Web Text))

Max Activating Token Index: 341

Text #17

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.2295. Min Act: -0.0000

Data Index: 518576 (C4 (Web Text))

Max Activating Token Index: 580

Text #18

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.0884. Min Act: -0.0001

Data Index: 434015 (C4 (Web Text))

Max Activating Token Index: 833

Text #19

Max Range: 1.3853. Min Range: -1.3853

Max Act: 1.0840. Min Act: -0.0001

Data Index: 415795 (C4 (Web Text))

Max Activating Token Index: 260