Model: SoLU Model: 6 Layers, 3072 Neurons per Layer

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

Neuron 1774 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-6l')

Text #0

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0482. Min Act: -0.0003

Data Index: 606633 (C4 (Web Text))

Max Activating Token Index: 122

Text #1

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0337. Min Act: -0.0003

Data Index: 285042 (C4 (Web Text))

Max Activating Token Index: 972

Text #2

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0324. Min Act: -0.0003

Data Index: 1021408 (C4 (Web Text))

Max Activating Token Index: 288

Text #3

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0302. Min Act: -0.0003

Data Index: 388513 (C4 (Web Text))

Max Activating Token Index: 578

Text #4

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0322. Min Act: -0.0003

Data Index: 844084 (C4 (Web Text))

Max Activating Token Index: 564

Text #5

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0351. Min Act: -0.0003

Data Index: 1005020 (C4 (Web Text))

Max Activating Token Index: 628

Text #6

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0315. Min Act: -0.0003

Data Index: 1583187 (Python Code)

Max Activating Token Index: 823

Text #7

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0301. Min Act: -0.0003

Data Index: 175970 (C4 (Web Text))

Max Activating Token Index: 46

Text #8

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0277. Min Act: -0.0003

Data Index: 52433 (C4 (Web Text))

Max Activating Token Index: 441

Text #9

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0287. Min Act: -0.0003

Data Index: 840011 (C4 (Web Text))

Max Activating Token Index: 167

Text #10

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0283. Min Act: -0.0003

Data Index: 1548707 (Python Code)

Max Activating Token Index: 925

Text #11

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0300. Min Act: -0.0003

Data Index: 1576583 (Python Code)

Max Activating Token Index: 900

Text #12

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0252. Min Act: -0.0003

Data Index: 96902 (C4 (Web Text))

Max Activating Token Index: 383

Text #13

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0234. Min Act: -0.0003

Data Index: 352792 (C4 (Web Text))

Max Activating Token Index: 897

Text #14

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0258. Min Act: -0.0003

Data Index: 460840 (C4 (Web Text))

Max Activating Token Index: 538

Text #15

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0269. Min Act: -0.0003

Data Index: 1446281 (Python Code)

Max Activating Token Index: 66

Text #16

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0259. Min Act: -0.0003

Data Index: 1241334 (C4 (Web Text))

Max Activating Token Index: 307

Text #17

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0276. Min Act: -0.0003

Data Index: 1038060 (C4 (Web Text))

Max Activating Token Index: 672

Text #18

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0249. Min Act: -0.0003

Data Index: 745201 (C4 (Web Text))

Max Activating Token Index: 21

Text #19

Max Range: 0.0482. Min Range: -0.0482

Max Act: 0.0260. Min Act: -0.0003

Data Index: 628003 (C4 (Web Text))

Max Activating Token Index: 553