Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 5785 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 3.5495. Min Range: -3.5495

Max Act: 3.5495. Min Act: -0.0000

Data Index: 170709 (C4 (Web Text))

Max Activating Token Index: 37

Text #1

Max Range: 3.5495. Min Range: -3.5495

Max Act: 3.0241. Min Act: -0.0000

Data Index: 325934 (C4 (Web Text))

Max Activating Token Index: 246

Text #2

Max Range: 3.5495. Min Range: -3.5495

Max Act: 3.0086. Min Act: -0.0000

Data Index: 1320413 (C4 (Web Text))

Max Activating Token Index: 217

Text #3

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.8141. Min Act: -0.0000

Data Index: 1593561 (Python Code)

Max Activating Token Index: 972

Text #4

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.8026. Min Act: -0.0000

Data Index: 760435 (C4 (Web Text))

Max Activating Token Index: 707

Text #5

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.7144. Min Act: -0.0000

Data Index: 181188 (C4 (Web Text))

Max Activating Token Index: 920

Text #6

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.8264. Min Act: -0.0000

Data Index: 589131 (C4 (Web Text))

Max Activating Token Index: 307

Text #7

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.6748. Min Act: -0.0000

Data Index: 988642 (C4 (Web Text))

Max Activating Token Index: 395

Text #8

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.5163. Min Act: -0.0000

Data Index: 1149482 (C4 (Web Text))

Max Activating Token Index: 158

Text #9

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.5866. Min Act: -0.0000

Data Index: 111901 (C4 (Web Text))

Max Activating Token Index: 477

Text #10

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.4615. Min Act: -0.0000

Data Index: 1190747 (C4 (Web Text))

Max Activating Token Index: 701

Text #11

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.5763. Min Act: -0.0000

Data Index: 1266234 (C4 (Web Text))

Max Activating Token Index: 142

Text #12

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.5261. Min Act: -0.0000

Data Index: 585502 (C4 (Web Text))

Max Activating Token Index: 890

Text #13

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.5234. Min Act: -0.0000

Data Index: 4940 (C4 (Web Text))

Max Activating Token Index: 406

Text #14

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.3482. Min Act: -0.0000

Data Index: 336474 (C4 (Web Text))

Max Activating Token Index: 870

Text #15

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.4757. Min Act: -0.0000

Data Index: 513838 (C4 (Web Text))

Max Activating Token Index: 786

Text #16

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.4543. Min Act: -0.0000

Data Index: 1440321 (Python Code)

Max Activating Token Index: 439

Text #17

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.4497. Min Act: -0.0000

Data Index: 81558 (C4 (Web Text))

Max Activating Token Index: 312

Text #18

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.4349. Min Act: -0.0000

Data Index: 636372 (C4 (Web Text))

Max Activating Token Index: 220

Text #19

Max Range: 3.5495. Min Range: -3.5495

Max Act: 2.4413. Min Act: -0.0000

Data Index: 804893 (C4 (Web Text))

Max Activating Token Index: 741