Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 1482 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0063. Min Act: -0.0006

Data Index: 586643 (C4 (Web Text))

Max Activating Token Index: 732

Text #1

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0063. Min Act: -0.0006

Data Index: 117254 (C4 (Web Text))

Max Activating Token Index: 891

Text #2

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0061. Min Act: -0.0006

Data Index: 746333 (C4 (Web Text))

Max Activating Token Index: 107

Text #3

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0058. Min Act: -0.0006

Data Index: 335823 (C4 (Web Text))

Max Activating Token Index: 852

Text #4

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0059. Min Act: -0.0006

Data Index: 1656589 (Python Code)

Max Activating Token Index: 167

Text #5

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0060. Min Act: -0.0006

Data Index: 1069160 (C4 (Web Text))

Max Activating Token Index: 606

Text #6

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0057. Min Act: -0.0006

Data Index: 1514469 (Python Code)

Max Activating Token Index: 472

Text #7

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0054. Min Act: -0.0006

Data Index: 1192749 (C4 (Web Text))

Max Activating Token Index: 890

Text #8

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0053. Min Act: -0.0006

Data Index: 1300191 (C4 (Web Text))

Max Activating Token Index: 586

Text #9

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0055. Min Act: -0.0006

Data Index: 1244963 (C4 (Web Text))

Max Activating Token Index: 47

Text #10

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0054. Min Act: -0.0006

Data Index: 1641570 (Python Code)

Max Activating Token Index: 440

Text #11

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0053. Min Act: -0.0006

Data Index: 257393 (C4 (Web Text))

Max Activating Token Index: 898

Text #12

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0054. Min Act: -0.0006

Data Index: 592952 (C4 (Web Text))

Max Activating Token Index: 585

Text #13

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0054. Min Act: -0.0006

Data Index: 188960 (C4 (Web Text))

Max Activating Token Index: 997

Text #14

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0052. Min Act: -0.0006

Data Index: 1097946 (C4 (Web Text))

Max Activating Token Index: 107

Text #15

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0055. Min Act: -0.0006

Data Index: 627442 (C4 (Web Text))

Max Activating Token Index: 970

Text #16

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0054. Min Act: -0.0006

Data Index: 387602 (C4 (Web Text))

Max Activating Token Index: 660

Text #17

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0053. Min Act: -0.0006

Data Index: 1359942 (Python Code)

Max Activating Token Index: 219

Text #18

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0051. Min Act: -0.0006

Data Index: 1512989 (Python Code)

Max Activating Token Index: 1003

Text #19

Max Range: 0.0063. Min Range: -0.0063

Max Act: 0.0052. Min Act: -0.0006

Data Index: 1557347 (Python Code)

Max Activating Token Index: 199