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

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

Neuron 3720 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 4.0497. Min Range: -4.0497

Max Act: 4.0497. Min Act: -0.0000

Data Index: 1006968 (C4 (Web Text))

Max Activating Token Index: 200

Text #1

Max Range: 4.0497. Min Range: -4.0497

Max Act: 3.5719. Min Act: -0.0000

Data Index: 454861 (C4 (Web Text))

Max Activating Token Index: 484

Text #2

Max Range: 4.0497. Min Range: -4.0497

Max Act: 3.5079. Min Act: -0.0000

Data Index: 1246115 (C4 (Web Text))

Max Activating Token Index: 220

Text #3

Max Range: 4.0497. Min Range: -4.0497

Max Act: 3.0336. Min Act: -0.0001

Data Index: 168559 (C4 (Web Text))

Max Activating Token Index: 836

Text #4

Max Range: 4.0497. Min Range: -4.0497

Max Act: 3.1539. Min Act: -0.0000

Data Index: 398512 (C4 (Web Text))

Max Activating Token Index: 37

Text #5

Max Range: 4.0497. Min Range: -4.0497

Max Act: 3.0715. Min Act: -0.0000

Data Index: 828179 (C4 (Web Text))

Max Activating Token Index: 252

Text #6

Max Range: 4.0497. Min Range: -4.0497

Max Act: 3.0251. Min Act: -0.0000

Data Index: 459633 (C4 (Web Text))

Max Activating Token Index: 144

Text #7

Max Range: 4.0497. Min Range: -4.0497

Max Act: 2.9943. Min Act: -0.0001

Data Index: 528941 (C4 (Web Text))

Max Activating Token Index: 932

Text #8

Max Range: 4.0497. Min Range: -4.0497

Max Act: 2.9932. Min Act: -0.0001

Data Index: 429092 (C4 (Web Text))

Max Activating Token Index: 125

Text #9

Max Range: 4.0497. Min Range: -4.0497

Max Act: 3.1299. Min Act: -0.0001

Data Index: 1333885 (C4 (Web Text))

Max Activating Token Index: 256

Text #10

Max Range: 4.0497. Min Range: -4.0497

Max Act: 2.9286. Min Act: -0.0001

Data Index: 560504 (C4 (Web Text))

Max Activating Token Index: 584

Text #11

Max Range: 4.0497. Min Range: -4.0497

Max Act: 2.8614. Min Act: -0.0001

Data Index: 802548 (C4 (Web Text))

Max Activating Token Index: 319

Text #12

Max Range: 4.0497. Min Range: -4.0497

Max Act: 2.8294. Min Act: -0.0001

Data Index: 381734 (C4 (Web Text))

Max Activating Token Index: 732

Text #13

Max Range: 4.0497. Min Range: -4.0497

Max Act: 2.8222. Min Act: -0.0001

Data Index: 937240 (C4 (Web Text))

Max Activating Token Index: 540

Text #14

Max Range: 4.0497. Min Range: -4.0497

Max Act: 2.7914. Min Act: -0.0001

Data Index: 113345 (C4 (Web Text))

Max Activating Token Index: 810

Text #15

Max Range: 4.0497. Min Range: -4.0497

Max Act: 2.7743. Min Act: -0.0000

Data Index: 144880 (C4 (Web Text))

Max Activating Token Index: 170

Text #16

Max Range: 4.0497. Min Range: -4.0497

Max Act: 2.7634. Min Act: -0.0001

Data Index: 873342 (C4 (Web Text))

Max Activating Token Index: 22

Text #17

Max Range: 4.0497. Min Range: -4.0497

Max Act: 2.7328. Min Act: -0.0001

Data Index: 38360 (C4 (Web Text))

Max Activating Token Index: 976

Text #18

Max Range: 4.0497. Min Range: -4.0497

Max Act: 2.6161. Min Act: -0.0001

Data Index: 332737 (C4 (Web Text))

Max Activating Token Index: 210

Text #19

Max Range: 4.0497. Min Range: -4.0497

Max Act: 2.5708. Min Act: -0.0000

Data Index: 323055 (C4 (Web Text))

Max Activating Token Index: 1007