Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 146 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1925. Min Act: -0.0001

Data Index: 360723 (C4 (Web Text))

Max Activating Token Index: 518

Text #1

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1666. Min Act: -0.0001

Data Index: 357581 (C4 (Web Text))

Max Activating Token Index: 63

Text #2

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1411. Min Act: -0.0001

Data Index: 235483 (C4 (Web Text))

Max Activating Token Index: 234

Text #3

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1357. Min Act: -0.0001

Data Index: 1036980 (C4 (Web Text))

Max Activating Token Index: 662

Text #4

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1342. Min Act: -0.0001

Data Index: 292234 (C4 (Web Text))

Max Activating Token Index: 244

Text #5

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1390. Min Act: -0.0001

Data Index: 1475928 (Python Code)

Max Activating Token Index: 887

Text #6

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1392. Min Act: -0.0001

Data Index: 1140185 (C4 (Web Text))

Max Activating Token Index: 342

Text #7

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1382. Min Act: -0.0001

Data Index: 739331 (C4 (Web Text))

Max Activating Token Index: 431

Text #8

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1369. Min Act: -0.0001

Data Index: 920578 (C4 (Web Text))

Max Activating Token Index: 97

Text #9

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1314. Min Act: -0.0001

Data Index: 714404 (C4 (Web Text))

Max Activating Token Index: 474

Text #10

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1457. Min Act: -0.0001

Data Index: 1102188 (C4 (Web Text))

Max Activating Token Index: 642

Text #11

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1347. Min Act: -0.0001

Data Index: 1365137 (Python Code)

Max Activating Token Index: 147

Text #12

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1346. Min Act: -0.0001

Data Index: 290051 (C4 (Web Text))

Max Activating Token Index: 242

Text #13

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1297. Min Act: -0.0001

Data Index: 887345 (C4 (Web Text))

Max Activating Token Index: 200

Text #14

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1350. Min Act: -0.0001

Data Index: 1086925 (C4 (Web Text))

Max Activating Token Index: 768

Text #15

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1197. Min Act: -0.0001

Data Index: 867414 (C4 (Web Text))

Max Activating Token Index: 966

Text #16

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1244. Min Act: -0.0001

Data Index: 765913 (C4 (Web Text))

Max Activating Token Index: 722

Text #17

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1279. Min Act: -0.0001

Data Index: 449167 (C4 (Web Text))

Max Activating Token Index: 596

Text #18

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1314. Min Act: -0.0001

Data Index: 1020106 (C4 (Web Text))

Max Activating Token Index: 511

Text #19

Max Range: 0.1925. Min Range: -0.1925

Max Act: 0.1166. Min Act: -0.0001

Data Index: 661779 (C4 (Web Text))

Max Activating Token Index: 88