Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 5271 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1190. Min Act: -0.0001

Data Index: 1334435 (C4 (Web Text))

Max Activating Token Index: 25

Text #1

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1253. Min Act: -0.0001

Data Index: 540734 (C4 (Web Text))

Max Activating Token Index: 119

Text #2

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1100. Min Act: -0.0001

Data Index: 161938 (C4 (Web Text))

Max Activating Token Index: 184

Text #3

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1134. Min Act: -0.0001

Data Index: 1176477 (C4 (Web Text))

Max Activating Token Index: 544

Text #4

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1142. Min Act: -0.0001

Data Index: 148090 (C4 (Web Text))

Max Activating Token Index: 981

Text #5

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1156. Min Act: -0.0001

Data Index: 562166 (C4 (Web Text))

Max Activating Token Index: 508

Text #6

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1138. Min Act: -0.0001

Data Index: 1313758 (C4 (Web Text))

Max Activating Token Index: 931

Text #7

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1037. Min Act: -0.0001

Data Index: 82319 (C4 (Web Text))

Max Activating Token Index: 292

Text #8

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1040. Min Act: -0.0001

Data Index: 513116 (C4 (Web Text))

Max Activating Token Index: 663

Text #9

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1211. Min Act: -0.0001

Data Index: 690745 (C4 (Web Text))

Max Activating Token Index: 145

Text #10

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1083. Min Act: -0.0001

Data Index: 1564873 (Python Code)

Max Activating Token Index: 431

Text #11

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1156. Min Act: -0.0001

Data Index: 655097 (C4 (Web Text))

Max Activating Token Index: 634

Text #12

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1157. Min Act: -0.0001

Data Index: 1133862 (C4 (Web Text))

Max Activating Token Index: 855

Text #13

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1115. Min Act: -0.0002

Data Index: 1073861 (C4 (Web Text))

Max Activating Token Index: 485

Text #14

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1077. Min Act: -0.0001

Data Index: 289859 (C4 (Web Text))

Max Activating Token Index: 900

Text #15

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1072. Min Act: -0.0001

Data Index: 680785 (C4 (Web Text))

Max Activating Token Index: 600

Text #16

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1113. Min Act: -0.0001

Data Index: 929021 (C4 (Web Text))

Max Activating Token Index: 576

Text #17

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1106. Min Act: -0.0001

Data Index: 378424 (C4 (Web Text))

Max Activating Token Index: 100

Text #18

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1065. Min Act: -0.0001

Data Index: 1443310 (Python Code)

Max Activating Token Index: 352

Text #19

Max Range: 0.1253. Min Range: -0.1253

Max Act: 0.1106. Min Act: -0.0001

Data Index: 1064833 (C4 (Web Text))

Max Activating Token Index: 39