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

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

Neuron 3834 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.7565. Min Act: -0.0001

Data Index: 1492950 (Python Code)

Max Activating Token Index: 437

Text #1

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.7017. Min Act: -0.0001

Data Index: 1394324 (Python Code)

Max Activating Token Index: 233

Text #2

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.6584. Min Act: -0.0001

Data Index: 1428129 (Python Code)

Max Activating Token Index: 456

Text #3

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5854. Min Act: -0.0001

Data Index: 1053906 (C4 (Web Text))

Max Activating Token Index: 801

Text #4

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.6095. Min Act: -0.0001

Data Index: 911552 (C4 (Web Text))

Max Activating Token Index: 710

Text #5

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5871. Min Act: -0.0001

Data Index: 397327 (C4 (Web Text))

Max Activating Token Index: 952

Text #6

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5991. Min Act: -0.0001

Data Index: 1479238 (Python Code)

Max Activating Token Index: 353

Text #7

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5807. Min Act: -0.0001

Data Index: 186782 (C4 (Web Text))

Max Activating Token Index: 971

Text #8

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5692. Min Act: -0.0001

Data Index: 1500731 (Python Code)

Max Activating Token Index: 439

Text #9

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5701. Min Act: -0.0001

Data Index: 460186 (C4 (Web Text))

Max Activating Token Index: 906

Text #10

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5469. Min Act: -0.0001

Data Index: 743089 (C4 (Web Text))

Max Activating Token Index: 80

Text #11

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5671. Min Act: -0.0001

Data Index: 1651487 (Python Code)

Max Activating Token Index: 449

Text #12

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5827. Min Act: -0.0001

Data Index: 631022 (C4 (Web Text))

Max Activating Token Index: 937

Text #13

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5471. Min Act: -0.0001

Data Index: 1036121 (C4 (Web Text))

Max Activating Token Index: 1001

Text #14

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5629. Min Act: -0.0001

Data Index: 454296 (C4 (Web Text))

Max Activating Token Index: 369

Text #15

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5623. Min Act: -0.0001

Data Index: 592865 (C4 (Web Text))

Max Activating Token Index: 994

Text #16

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5546. Min Act: -0.0001

Data Index: 1513025 (Python Code)

Max Activating Token Index: 322

Text #17

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5590. Min Act: -0.0001

Data Index: 696897 (C4 (Web Text))

Max Activating Token Index: 986

Text #18

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5529. Min Act: -0.0001

Data Index: 1412862 (Python Code)

Max Activating Token Index: 444

Text #19

Max Range: 0.7565. Min Range: -0.7565

Max Act: 0.5888. Min Act: -0.0001

Data Index: 1093214 (C4 (Web Text))

Max Activating Token Index: 927