Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 1796 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.5485. Min Act: -0.0001

Data Index: 1574094 (Python Code)

Max Activating Token Index: 743

Text #1

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.5299. Min Act: 0.0000

Data Index: 1453902 (Python Code)

Max Activating Token Index: 958

Text #2

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.4680. Min Act: -0.0001

Data Index: 84918 (C4 (Web Text))

Max Activating Token Index: 1013

Text #3

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.4595. Min Act: 0.0000

Data Index: 1404777 (Python Code)

Max Activating Token Index: 271

Text #4

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.4564. Min Act: 0.0000

Data Index: 168987 (C4 (Web Text))

Max Activating Token Index: 736

Text #5

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.4547. Min Act: 0.0000

Data Index: 1552289 (Python Code)

Max Activating Token Index: 739

Text #6

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.4099. Min Act: 0.0000

Data Index: 1427417 (Python Code)

Max Activating Token Index: 773

Text #7

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.4052. Min Act: -0.0001

Data Index: 1046226 (C4 (Web Text))

Max Activating Token Index: 447

Text #8

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.4150. Min Act: -0.0001

Data Index: 1654821 (Python Code)

Max Activating Token Index: 861

Text #9

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.3892. Min Act: -0.0000

Data Index: 1512453 (Python Code)

Max Activating Token Index: 385

Text #10

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.3676. Min Act: -0.0002

Data Index: 93097 (C4 (Web Text))

Max Activating Token Index: 302

Text #11

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.3728. Min Act: -0.0001

Data Index: 247264 (C4 (Web Text))

Max Activating Token Index: 331

Text #12

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.3985. Min Act: 0.0000

Data Index: 1627534 (Python Code)

Max Activating Token Index: 764

Text #13

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.3553. Min Act: 0.0000

Data Index: 1403450 (Python Code)

Max Activating Token Index: 814

Text #14

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.3678. Min Act: -0.0001

Data Index: 1435265 (Python Code)

Max Activating Token Index: 1004

Text #15

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.3666. Min Act: 0.0000

Data Index: 1645224 (Python Code)

Max Activating Token Index: 894

Text #16

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.3489. Min Act: -0.0001

Data Index: 975826 (C4 (Web Text))

Max Activating Token Index: 583

Text #17

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.3606. Min Act: -0.0000

Data Index: 1538629 (Python Code)

Max Activating Token Index: 26

Text #18

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.3559. Min Act: 0.0000

Data Index: 1630861 (Python Code)

Max Activating Token Index: 869

Text #19

Max Range: 0.5485. Min Range: -0.5485

Max Act: 0.3664. Min Act: -0.0001

Data Index: 1455663 (Python Code)

Max Activating Token Index: 899