Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 224 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0757. Min Act: -0.0002

Data Index: 296807 (C4 (Web Text))

Max Activating Token Index: 806

Text #1

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0754. Min Act: -0.0003

Data Index: 1557301 (Python Code)

Max Activating Token Index: 544

Text #2

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0751. Min Act: -0.0003

Data Index: 9032 (C4 (Web Text))

Max Activating Token Index: 217

Text #3

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0747. Min Act: -0.0003

Data Index: 485617 (C4 (Web Text))

Max Activating Token Index: 523

Text #4

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0715. Min Act: -0.0003

Data Index: 1285928 (C4 (Web Text))

Max Activating Token Index: 329

Text #5

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0706. Min Act: -0.0003

Data Index: 169821 (C4 (Web Text))

Max Activating Token Index: 973

Text #6

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0699. Min Act: -0.0003

Data Index: 1107252 (C4 (Web Text))

Max Activating Token Index: 1012

Text #7

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0693. Min Act: -0.0003

Data Index: 1577428 (Python Code)

Max Activating Token Index: 762

Text #8

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0701. Min Act: -0.0003

Data Index: 1307497 (C4 (Web Text))

Max Activating Token Index: 161

Text #9

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0708. Min Act: -0.0002

Data Index: 741245 (C4 (Web Text))

Max Activating Token Index: 436

Text #10

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0709. Min Act: -0.0003

Data Index: 1314175 (C4 (Web Text))

Max Activating Token Index: 406

Text #11

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0692. Min Act: -0.0004

Data Index: 76792 (C4 (Web Text))

Max Activating Token Index: 436

Text #12

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0701. Min Act: -0.0003

Data Index: 835137 (C4 (Web Text))

Max Activating Token Index: 466

Text #13

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0689. Min Act: -0.0003

Data Index: 342258 (C4 (Web Text))

Max Activating Token Index: 275

Text #14

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0674. Min Act: -0.0002

Data Index: 1122403 (C4 (Web Text))

Max Activating Token Index: 392

Text #15

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0672. Min Act: -0.0003

Data Index: 11134 (C4 (Web Text))

Max Activating Token Index: 1018

Text #16

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0676. Min Act: -0.0003

Data Index: 848962 (C4 (Web Text))

Max Activating Token Index: 40

Text #17

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0665. Min Act: -0.0002

Data Index: 130531 (C4 (Web Text))

Max Activating Token Index: 775

Text #18

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0657. Min Act: -0.0003

Data Index: 1155383 (C4 (Web Text))

Max Activating Token Index: 613

Text #19

Max Range: 0.0757. Min Range: -0.0757

Max Act: 0.0662. Min Act: -0.0003

Data Index: 957120 (C4 (Web Text))

Max Activating Token Index: 676