Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 2011 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0854. Min Act: -0.0006

Data Index: 1620448 (Python Code)

Max Activating Token Index: 156

Text #1

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0364. Min Act: -0.0006

Data Index: 1204826 (C4 (Web Text))

Max Activating Token Index: 938

Text #2

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0362. Min Act: -0.0005

Data Index: 1469663 (Python Code)

Max Activating Token Index: 1011

Text #3

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0356. Min Act: -0.0006

Data Index: 366874 (C4 (Web Text))

Max Activating Token Index: 949

Text #4

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0348. Min Act: -0.0006

Data Index: 988046 (C4 (Web Text))

Max Activating Token Index: 716

Text #5

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0331. Min Act: -0.0006

Data Index: 1222561 (C4 (Web Text))

Max Activating Token Index: 740

Text #6

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0310. Min Act: -0.0006

Data Index: 853913 (C4 (Web Text))

Max Activating Token Index: 957

Text #7

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0296. Min Act: -0.0006

Data Index: 315934 (C4 (Web Text))

Max Activating Token Index: 997

Text #8

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0288. Min Act: -0.0006

Data Index: 1451842 (Python Code)

Max Activating Token Index: 827

Text #9

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0290. Min Act: -0.0006

Data Index: 276593 (C4 (Web Text))

Max Activating Token Index: 889

Text #10

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0268. Min Act: -0.0006

Data Index: 650453 (C4 (Web Text))

Max Activating Token Index: 618

Text #11

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0278. Min Act: -0.0006

Data Index: 1571680 (Python Code)

Max Activating Token Index: 572

Text #12

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0267. Min Act: -0.0006

Data Index: 1155971 (C4 (Web Text))

Max Activating Token Index: 685

Text #13

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0269. Min Act: -0.0006

Data Index: 1516650 (Python Code)

Max Activating Token Index: 363

Text #14

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0269. Min Act: -0.0006

Data Index: 630617 (C4 (Web Text))

Max Activating Token Index: 469

Text #15

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0267. Min Act: -0.0006

Data Index: 181529 (C4 (Web Text))

Max Activating Token Index: 873

Text #16

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0269. Min Act: -0.0005

Data Index: 1546648 (Python Code)

Max Activating Token Index: 591

Text #17

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0262. Min Act: -0.0006

Data Index: 790458 (C4 (Web Text))

Max Activating Token Index: 984

Text #18

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0260. Min Act: -0.0006

Data Index: 334928 (C4 (Web Text))

Max Activating Token Index: 627

Text #19

Max Range: 0.0854. Min Range: -0.0854

Max Act: 0.0264. Min Act: -0.0006

Data Index: 1651595 (Python Code)

Max Activating Token Index: 732