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

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

Neuron 3949 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1246. Min Act: -0.0002

Data Index: 1645076 (Python Code)

Max Activating Token Index: 528

Text #1

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1086. Min Act: -0.0002

Data Index: 1119382 (C4 (Web Text))

Max Activating Token Index: 109

Text #2

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1023. Min Act: -0.0002

Data Index: 328526 (C4 (Web Text))

Max Activating Token Index: 661

Text #3

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1011. Min Act: -0.0002

Data Index: 999012 (C4 (Web Text))

Max Activating Token Index: 776

Text #4

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.0970. Min Act: -0.0002

Data Index: 1215803 (C4 (Web Text))

Max Activating Token Index: 172

Text #5

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.0987. Min Act: -0.0002

Data Index: 110847 (C4 (Web Text))

Max Activating Token Index: 824

Text #6

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1278. Min Act: -0.0002

Data Index: 534330 (C4 (Web Text))

Max Activating Token Index: 535

Text #7

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1006. Min Act: -0.0002

Data Index: 127694 (C4 (Web Text))

Max Activating Token Index: 224

Text #8

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1031. Min Act: -0.0002

Data Index: 170047 (C4 (Web Text))

Max Activating Token Index: 20

Text #9

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1111. Min Act: -0.0002

Data Index: 465129 (C4 (Web Text))

Max Activating Token Index: 926

Text #10

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1052. Min Act: -0.0002

Data Index: 1099299 (C4 (Web Text))

Max Activating Token Index: 419

Text #11

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1082. Min Act: -0.0002

Data Index: 239635 (C4 (Web Text))

Max Activating Token Index: 622

Text #12

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1097. Min Act: -0.0002

Data Index: 1221821 (C4 (Web Text))

Max Activating Token Index: 163

Text #13

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1124. Min Act: -0.0002

Data Index: 1574017 (Python Code)

Max Activating Token Index: 399

Text #14

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1129. Min Act: -0.0002

Data Index: 433547 (C4 (Web Text))

Max Activating Token Index: 702

Text #15

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1093. Min Act: -0.0002

Data Index: 1606694 (Python Code)

Max Activating Token Index: 431

Text #16

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.0971. Min Act: -0.0002

Data Index: 721359 (C4 (Web Text))

Max Activating Token Index: 67

Text #17

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.1258. Min Act: -0.0002

Data Index: 381334 (C4 (Web Text))

Max Activating Token Index: 516

Text #18

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.0965. Min Act: -0.0002

Data Index: 542517 (C4 (Web Text))

Max Activating Token Index: 81

Text #19

Max Range: 0.1278. Min Range: -0.1278

Max Act: 0.0998. Min Act: -0.0002

Data Index: 654657 (C4 (Web Text))

Max Activating Token Index: 741