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

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

Neuron 3648 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2624. Min Act: -0.0000

Data Index: 1101582 (C4 (Web Text))

Max Activating Token Index: 206

Text #1

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2177. Min Act: -0.0000

Data Index: 940211 (C4 (Web Text))

Max Activating Token Index: 243

Text #2

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2652. Min Act: -0.0000

Data Index: 942654 (C4 (Web Text))

Max Activating Token Index: 363

Text #3

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2195. Min Act: -0.0000

Data Index: 837753 (C4 (Web Text))

Max Activating Token Index: 315

Text #4

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2278. Min Act: -0.0001

Data Index: 1170035 (C4 (Web Text))

Max Activating Token Index: 782

Text #5

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2219. Min Act: -0.0001

Data Index: 1169483 (C4 (Web Text))

Max Activating Token Index: 130

Text #6

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2249. Min Act: -0.0001

Data Index: 1125814 (C4 (Web Text))

Max Activating Token Index: 23

Text #7

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2078. Min Act: -0.0000

Data Index: 1175999 (C4 (Web Text))

Max Activating Token Index: 572

Text #8

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2218. Min Act: -0.0001

Data Index: 450597 (C4 (Web Text))

Max Activating Token Index: 190

Text #9

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.1984. Min Act: -0.0000

Data Index: 758431 (C4 (Web Text))

Max Activating Token Index: 524

Text #10

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2202. Min Act: -0.0000

Data Index: 1313174 (C4 (Web Text))

Max Activating Token Index: 817

Text #11

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2003. Min Act: -0.0001

Data Index: 548271 (C4 (Web Text))

Max Activating Token Index: 896

Text #12

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.1948. Min Act: -0.0001

Data Index: 1405564 (Python Code)

Max Activating Token Index: 590

Text #13

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2087. Min Act: -0.0000

Data Index: 915905 (C4 (Web Text))

Max Activating Token Index: 552

Text #14

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2030. Min Act: -0.0000

Data Index: 938348 (C4 (Web Text))

Max Activating Token Index: 223

Text #15

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2053. Min Act: -0.0001

Data Index: 606470 (C4 (Web Text))

Max Activating Token Index: 15

Text #16

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2184. Min Act: -0.0000

Data Index: 547559 (C4 (Web Text))

Max Activating Token Index: 900

Text #17

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2113. Min Act: -0.0001

Data Index: 959716 (C4 (Web Text))

Max Activating Token Index: 820

Text #18

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.1999. Min Act: -0.0001

Data Index: 676710 (C4 (Web Text))

Max Activating Token Index: 256

Text #19

Max Range: 0.2652. Min Range: -0.2652

Max Act: 0.2011. Min Act: -0.0000

Data Index: 696491 (C4 (Web Text))

Max Activating Token Index: 383