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

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

Neuron 259 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: 0.5230. Min Range: -0.5230

Max Act: 0.5230. Min Act: -0.0001

Data Index: 1296655 (C4 (Web Text))

Max Activating Token Index: 963

Text #1

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4862. Min Act: -0.0001

Data Index: 717276 (C4 (Web Text))

Max Activating Token Index: 873

Text #2

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4519. Min Act: -0.0001

Data Index: 1641526 (Python Code)

Max Activating Token Index: 281

Text #3

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4926. Min Act: -0.0001

Data Index: 1481586 (Python Code)

Max Activating Token Index: 643

Text #4

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4790. Min Act: -0.0001

Data Index: 217471 (C4 (Web Text))

Max Activating Token Index: 102

Text #5

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4616. Min Act: -0.0001

Data Index: 1396910 (Python Code)

Max Activating Token Index: 421

Text #6

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4413. Min Act: -0.0001

Data Index: 164323 (C4 (Web Text))

Max Activating Token Index: 910

Text #7

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4363. Min Act: -0.0001

Data Index: 60613 (C4 (Web Text))

Max Activating Token Index: 756

Text #8

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4326. Min Act: -0.0001

Data Index: 1413475 (Python Code)

Max Activating Token Index: 24

Text #9

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4351. Min Act: -0.0001

Data Index: 1425423 (Python Code)

Max Activating Token Index: 230

Text #10

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4065. Min Act: -0.0001

Data Index: 1539082 (Python Code)

Max Activating Token Index: 148

Text #11

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4004. Min Act: -0.0001

Data Index: 1639371 (Python Code)

Max Activating Token Index: 122

Text #12

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4171. Min Act: -0.0001

Data Index: 515900 (C4 (Web Text))

Max Activating Token Index: 480

Text #13

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4000. Min Act: -0.0001

Data Index: 1481092 (Python Code)

Max Activating Token Index: 753

Text #14

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4221. Min Act: -0.0001

Data Index: 1154860 (C4 (Web Text))

Max Activating Token Index: 182

Text #15

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4209. Min Act: -0.0001

Data Index: 1389327 (Python Code)

Max Activating Token Index: 37

Text #16

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4153. Min Act: -0.0001

Data Index: 1481963 (Python Code)

Max Activating Token Index: 711

Text #17

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4169. Min Act: -0.0001

Data Index: 84617 (C4 (Web Text))

Max Activating Token Index: 596

Text #18

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.4060. Min Act: -0.0001

Data Index: 1445227 (Python Code)

Max Activating Token Index: 661

Text #19

Max Range: 0.5230. Min Range: -0.5230

Max Act: 0.3913. Min Act: -0.0001

Data Index: 1401347 (Python Code)

Max Activating Token Index: 636