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

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

Neuron 4663 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1392. Min Act: -0.0001

Data Index: 740122 (C4 (Web Text))

Max Activating Token Index: 336

Text #1

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1197. Min Act: -0.0001

Data Index: 1409226 (Python Code)

Max Activating Token Index: 168

Text #2

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1260. Min Act: -0.0001

Data Index: 1142961 (C4 (Web Text))

Max Activating Token Index: 777

Text #3

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1260. Min Act: -0.0001

Data Index: 1146423 (C4 (Web Text))

Max Activating Token Index: 1008

Text #4

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1318. Min Act: -0.0001

Data Index: 626855 (C4 (Web Text))

Max Activating Token Index: 289

Text #5

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1240. Min Act: -0.0001

Data Index: 114531 (C4 (Web Text))

Max Activating Token Index: 93

Text #6

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1249. Min Act: -0.0001

Data Index: 1388296 (Python Code)

Max Activating Token Index: 632

Text #7

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1219. Min Act: -0.0001

Data Index: 840040 (C4 (Web Text))

Max Activating Token Index: 325

Text #8

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1003. Min Act: -0.0001

Data Index: 1277284 (C4 (Web Text))

Max Activating Token Index: 608

Text #9

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1171. Min Act: -0.0001

Data Index: 796141 (C4 (Web Text))

Max Activating Token Index: 467

Text #10

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1110. Min Act: -0.0001

Data Index: 169808 (C4 (Web Text))

Max Activating Token Index: 149

Text #11

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1126. Min Act: -0.0001

Data Index: 290504 (C4 (Web Text))

Max Activating Token Index: 47

Text #12

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1120. Min Act: -0.0001

Data Index: 419837 (C4 (Web Text))

Max Activating Token Index: 406

Text #13

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1112. Min Act: -0.0001

Data Index: 1134017 (C4 (Web Text))

Max Activating Token Index: 24

Text #14

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1025. Min Act: -0.0001

Data Index: 356067 (C4 (Web Text))

Max Activating Token Index: 200

Text #15

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1059. Min Act: -0.0001

Data Index: 606805 (C4 (Web Text))

Max Activating Token Index: 712

Text #16

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1125. Min Act: -0.0001

Data Index: 1044381 (C4 (Web Text))

Max Activating Token Index: 250

Text #17

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1092. Min Act: -0.0001

Data Index: 1320166 (C4 (Web Text))

Max Activating Token Index: 753

Text #18

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1095. Min Act: -0.0001

Data Index: 768245 (C4 (Web Text))

Max Activating Token Index: 385

Text #19

Max Range: 0.1392. Min Range: -0.1392

Max Act: 0.1057. Min Act: -0.0001

Data Index: 85273 (C4 (Web Text))

Max Activating Token Index: 12