Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

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

Neuron 3333 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.5123. Min Act: -0.0000

Data Index: 1247382 (C4 (Web Text))

Max Activating Token Index: 359

Text #1

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.4461. Min Act: -0.0000

Data Index: 1029301 (C4 (Web Text))

Max Activating Token Index: 999

Text #2

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.3473. Min Act: -0.0000

Data Index: 916714 (C4 (Web Text))

Max Activating Token Index: 96

Text #3

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2659. Min Act: -0.0000

Data Index: 512753 (C4 (Web Text))

Max Activating Token Index: 710

Text #4

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2429. Min Act: -0.0000

Data Index: 697802 (C4 (Web Text))

Max Activating Token Index: 665

Text #5

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2730. Min Act: -0.0000

Data Index: 336944 (C4 (Web Text))

Max Activating Token Index: 969

Text #6

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2569. Min Act: -0.0000

Data Index: 1292788 (C4 (Web Text))

Max Activating Token Index: 783

Text #7

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2274. Min Act: -0.0000

Data Index: 243879 (C4 (Web Text))

Max Activating Token Index: 360

Text #8

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2511. Min Act: -0.0001

Data Index: 664142 (C4 (Web Text))

Max Activating Token Index: 239

Text #9

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2078. Min Act: -0.0000

Data Index: 599893 (C4 (Web Text))

Max Activating Token Index: 835

Text #10

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2474. Min Act: -0.0000

Data Index: 265959 (C4 (Web Text))

Max Activating Token Index: 223

Text #11

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2232. Min Act: -0.0000

Data Index: 1078388 (C4 (Web Text))

Max Activating Token Index: 129

Text #12

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2174. Min Act: -0.0000

Data Index: 1314011 (C4 (Web Text))

Max Activating Token Index: 625

Text #13

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2263. Min Act: -0.0000

Data Index: 1478416 (Python Code)

Max Activating Token Index: 979

Text #14

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2130. Min Act: -0.0000

Data Index: 534161 (C4 (Web Text))

Max Activating Token Index: 681

Text #15

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2127. Min Act: -0.0000

Data Index: 1598646 (Python Code)

Max Activating Token Index: 527

Text #16

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2309. Min Act: -0.0001

Data Index: 1353868 (C4 (Web Text))

Max Activating Token Index: 857

Text #17

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2319. Min Act: -0.0000

Data Index: 328772 (C4 (Web Text))

Max Activating Token Index: 210

Text #18

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.2132. Min Act: -0.0000

Data Index: 673638 (C4 (Web Text))

Max Activating Token Index: 1006

Text #19

Max Range: 0.5123. Min Range: -0.5123

Max Act: 0.1947. Min Act: -0.0000

Data Index: 521055 (C4 (Web Text))

Max Activating Token Index: 338