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

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

Neuron 544 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1546. Min Act: -0.0000

Data Index: 42688 (C4 (Web Text))

Max Activating Token Index: 93

Text #1

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1389. Min Act: -0.0000

Data Index: 524285 (C4 (Web Text))

Max Activating Token Index: 566

Text #2

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1242. Min Act: -0.0000

Data Index: 500262 (C4 (Web Text))

Max Activating Token Index: 266

Text #3

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1268. Min Act: -0.0000

Data Index: 1051249 (C4 (Web Text))

Max Activating Token Index: 978

Text #4

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1196. Min Act: -0.0000

Data Index: 1314363 (C4 (Web Text))

Max Activating Token Index: 852

Text #5

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1109. Min Act: -0.0001

Data Index: 610536 (C4 (Web Text))

Max Activating Token Index: 271

Text #6

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1171. Min Act: -0.0001

Data Index: 161541 (C4 (Web Text))

Max Activating Token Index: 871

Text #7

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1120. Min Act: -0.0000

Data Index: 229040 (C4 (Web Text))

Max Activating Token Index: 721

Text #8

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1089. Min Act: -0.0000

Data Index: 631163 (C4 (Web Text))

Max Activating Token Index: 774

Text #9

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1135. Min Act: -0.0000

Data Index: 98503 (C4 (Web Text))

Max Activating Token Index: 717

Text #10

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1092. Min Act: -0.0001

Data Index: 368695 (C4 (Web Text))

Max Activating Token Index: 476

Text #11

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1088. Min Act: -0.0000

Data Index: 409114 (C4 (Web Text))

Max Activating Token Index: 682

Text #12

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1087. Min Act: -0.0000

Data Index: 515444 (C4 (Web Text))

Max Activating Token Index: 546

Text #13

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1045. Min Act: -0.0000

Data Index: 209976 (C4 (Web Text))

Max Activating Token Index: 567

Text #14

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1079. Min Act: -0.0000

Data Index: 838067 (C4 (Web Text))

Max Activating Token Index: 77

Text #15

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1059. Min Act: -0.0000

Data Index: 38549 (C4 (Web Text))

Max Activating Token Index: 667

Text #16

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1010. Min Act: -0.0001

Data Index: 1539303 (Python Code)

Max Activating Token Index: 864

Text #17

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1082. Min Act: -0.0000

Data Index: 1282086 (C4 (Web Text))

Max Activating Token Index: 795

Text #18

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1040. Min Act: -0.0000

Data Index: 681176 (C4 (Web Text))

Max Activating Token Index: 128

Text #19

Max Range: 0.1546. Min Range: -0.1546

Max Act: 0.1030. Min Act: -0.0001

Data Index: 888594 (C4 (Web Text))

Max Activating Token Index: 106