Model: SoLU Model: 6 Layers, 3072 Neurons per Layer

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

Neuron 1833 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-6l')

Text #0

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1539. Min Act: -0.0001

Data Index: 355885 (C4 (Web Text))

Max Activating Token Index: 104

Text #1

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1393. Min Act: -0.0001

Data Index: 573038 (C4 (Web Text))

Max Activating Token Index: 275

Text #2

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1277. Min Act: -0.0001

Data Index: 487315 (C4 (Web Text))

Max Activating Token Index: 207

Text #3

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1276. Min Act: -0.0001

Data Index: 395191 (C4 (Web Text))

Max Activating Token Index: 945

Text #4

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1209. Min Act: -0.0001

Data Index: 814609 (C4 (Web Text))

Max Activating Token Index: 297

Text #5

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1247. Min Act: -0.0001

Data Index: 1544464 (Python Code)

Max Activating Token Index: 100

Text #6

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1145. Min Act: -0.0001

Data Index: 6800 (C4 (Web Text))

Max Activating Token Index: 299

Text #7

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1176. Min Act: -0.0001

Data Index: 207608 (C4 (Web Text))

Max Activating Token Index: 169

Text #8

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1168. Min Act: -0.0001

Data Index: 1502187 (Python Code)

Max Activating Token Index: 509

Text #9

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1228. Min Act: -0.0001

Data Index: 1567357 (Python Code)

Max Activating Token Index: 802

Text #10

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1122. Min Act: -0.0001

Data Index: 272677 (C4 (Web Text))

Max Activating Token Index: 897

Text #11

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1190. Min Act: -0.0001

Data Index: 1609342 (Python Code)

Max Activating Token Index: 919

Text #12

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1107. Min Act: -0.0001

Data Index: 1371064 (Python Code)

Max Activating Token Index: 516

Text #13

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1093. Min Act: -0.0001

Data Index: 1234982 (C4 (Web Text))

Max Activating Token Index: 982

Text #14

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1046. Min Act: -0.0001

Data Index: 1376379 (Python Code)

Max Activating Token Index: 997

Text #15

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1012. Min Act: -0.0001

Data Index: 1198539 (C4 (Web Text))

Max Activating Token Index: 974

Text #16

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1091. Min Act: -0.0001

Data Index: 1547240 (Python Code)

Max Activating Token Index: 688

Text #17

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1158. Min Act: -0.0001

Data Index: 558703 (C4 (Web Text))

Max Activating Token Index: 981

Text #18

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1153. Min Act: -0.0001

Data Index: 1551396 (Python Code)

Max Activating Token Index: 887

Text #19

Max Range: 0.1539. Min Range: -0.1539

Max Act: 0.1099. Min Act: -0.0001

Data Index: 873827 (C4 (Web Text))

Max Activating Token Index: 586