Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 315 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0468. Min Act: -0.0005

Data Index: 1129209 (C4 (Web Text))

Max Activating Token Index: 739

Text #1

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0388. Min Act: -0.0005

Data Index: 787080 (C4 (Web Text))

Max Activating Token Index: 56

Text #2

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0337. Min Act: -0.0005

Data Index: 979405 (C4 (Web Text))

Max Activating Token Index: 62

Text #3

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0363. Min Act: -0.0005

Data Index: 31970 (C4 (Web Text))

Max Activating Token Index: 863

Text #4

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0338. Min Act: -0.0005

Data Index: 1255946 (C4 (Web Text))

Max Activating Token Index: 702

Text #5

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0345. Min Act: -0.0005

Data Index: 509378 (C4 (Web Text))

Max Activating Token Index: 84

Text #6

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0331. Min Act: -0.0005

Data Index: 341382 (C4 (Web Text))

Max Activating Token Index: 434

Text #7

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0347. Min Act: -0.0005

Data Index: 26526 (C4 (Web Text))

Max Activating Token Index: 707

Text #8

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0333. Min Act: -0.0005

Data Index: 515772 (C4 (Web Text))

Max Activating Token Index: 1023

Text #9

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0340. Min Act: -0.0005

Data Index: 182756 (C4 (Web Text))

Max Activating Token Index: 280

Text #10

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0326. Min Act: -0.0005

Data Index: 865215 (C4 (Web Text))

Max Activating Token Index: 607

Text #11

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0280. Min Act: -0.0005

Data Index: 581262 (C4 (Web Text))

Max Activating Token Index: 494

Text #12

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0309. Min Act: -0.0005

Data Index: 381522 (C4 (Web Text))

Max Activating Token Index: 19

Text #13

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0289. Min Act: -0.0005

Data Index: 826949 (C4 (Web Text))

Max Activating Token Index: 898

Text #14

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0322. Min Act: -0.0005

Data Index: 279248 (C4 (Web Text))

Max Activating Token Index: 756

Text #15

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0308. Min Act: -0.0005

Data Index: 335521 (C4 (Web Text))

Max Activating Token Index: 46

Text #16

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0284. Min Act: -0.0005

Data Index: 351002 (C4 (Web Text))

Max Activating Token Index: 672

Text #17

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0295. Min Act: -0.0005

Data Index: 945109 (C4 (Web Text))

Max Activating Token Index: 384

Text #18

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0319. Min Act: -0.0005

Data Index: 1509500 (Python Code)

Max Activating Token Index: 114

Text #19

Max Range: 0.0468. Min Range: -0.0468

Max Act: 0.0315. Min Act: -0.0005

Data Index: 38624 (C4 (Web Text))

Max Activating Token Index: 643