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

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

Neuron 2750 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.3065. Min Act: -0.0000

Data Index: 116375 (C4 (Web Text))

Max Activating Token Index: 505

Text #1

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2617. Min Act: -0.0000

Data Index: 1525114 (Python Code)

Max Activating Token Index: 471

Text #2

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2664. Min Act: -0.0000

Data Index: 17643 (C4 (Web Text))

Max Activating Token Index: 284

Text #3

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2815. Min Act: -0.0000

Data Index: 1369495 (Python Code)

Max Activating Token Index: 141

Text #4

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2544. Min Act: 0.0000

Data Index: 1437390 (Python Code)

Max Activating Token Index: 116

Text #5

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2404. Min Act: -0.0000

Data Index: 1611145 (Python Code)

Max Activating Token Index: 393

Text #6

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2348. Min Act: -0.0000

Data Index: 1395178 (Python Code)

Max Activating Token Index: 924

Text #7

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2355. Min Act: -0.0000

Data Index: 1379936 (Python Code)

Max Activating Token Index: 807

Text #8

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2534. Min Act: -0.0000

Data Index: 881172 (C4 (Web Text))

Max Activating Token Index: 705

Text #9

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2318. Min Act: -0.0000

Data Index: 1514158 (Python Code)

Max Activating Token Index: 521

Text #10

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2225. Min Act: -0.0000

Data Index: 1473701 (Python Code)

Max Activating Token Index: 475

Text #11

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2226. Min Act: -0.0000

Data Index: 102039 (C4 (Web Text))

Max Activating Token Index: 488

Text #12

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2168. Min Act: -0.0000

Data Index: 1506159 (Python Code)

Max Activating Token Index: 380

Text #13

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2004. Min Act: -0.0000

Data Index: 1377192 (Python Code)

Max Activating Token Index: 104

Text #14

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2043. Min Act: -0.0000

Data Index: 1481007 (Python Code)

Max Activating Token Index: 44

Text #15

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.1902. Min Act: -0.0000

Data Index: 155474 (C4 (Web Text))

Max Activating Token Index: 961

Text #16

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.1891. Min Act: -0.0000

Data Index: 1474004 (Python Code)

Max Activating Token Index: 368

Text #17

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2168. Min Act: -0.0000

Data Index: 910339 (C4 (Web Text))

Max Activating Token Index: 992

Text #18

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.2130. Min Act: -0.0000

Data Index: 605089 (C4 (Web Text))

Max Activating Token Index: 925

Text #19

Max Range: 0.3065. Min Range: -0.3065

Max Act: 0.1956. Min Act: -0.0000

Data Index: 1524674 (Python Code)

Max Activating Token Index: 118