Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 3728 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0867. Min Act: -0.0002

Data Index: 1014704 (C4 (Web Text))

Max Activating Token Index: 612

Text #1

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0826. Min Act: -0.0002

Data Index: 1058479 (C4 (Web Text))

Max Activating Token Index: 987

Text #2

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0757. Min Act: -0.0001

Data Index: 18616 (C4 (Web Text))

Max Activating Token Index: 925

Text #3

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0686. Min Act: -0.0002

Data Index: 1058655 (C4 (Web Text))

Max Activating Token Index: 214

Text #4

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0701. Min Act: -0.0002

Data Index: 703663 (C4 (Web Text))

Max Activating Token Index: 796

Text #5

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0681. Min Act: -0.0002

Data Index: 374870 (C4 (Web Text))

Max Activating Token Index: 629

Text #6

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0623. Min Act: -0.0001

Data Index: 181280 (C4 (Web Text))

Max Activating Token Index: 144

Text #7

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0649. Min Act: -0.0002

Data Index: 220964 (C4 (Web Text))

Max Activating Token Index: 24

Text #8

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0585. Min Act: -0.0002

Data Index: 325536 (C4 (Web Text))

Max Activating Token Index: 242

Text #9

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0570. Min Act: -0.0001

Data Index: 1441439 (Python Code)

Max Activating Token Index: 456

Text #10

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0618. Min Act: -0.0002

Data Index: 892728 (C4 (Web Text))

Max Activating Token Index: 618

Text #11

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0609. Min Act: -0.0001

Data Index: 173323 (C4 (Web Text))

Max Activating Token Index: 68

Text #12

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0608. Min Act: -0.0002

Data Index: 1206939 (C4 (Web Text))

Max Activating Token Index: 269

Text #13

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0555. Min Act: -0.0001

Data Index: 683276 (C4 (Web Text))

Max Activating Token Index: 579

Text #14

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0527. Min Act: -0.0001

Data Index: 1477887 (Python Code)

Max Activating Token Index: 45

Text #15

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0546. Min Act: -0.0001

Data Index: 149073 (C4 (Web Text))

Max Activating Token Index: 58

Text #16

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0557. Min Act: -0.0002

Data Index: 31137 (C4 (Web Text))

Max Activating Token Index: 894

Text #17

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0592. Min Act: -0.0001

Data Index: 1563984 (Python Code)

Max Activating Token Index: 298

Text #18

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0530. Min Act: -0.0001

Data Index: 1411032 (Python Code)

Max Activating Token Index: 180

Text #19

Max Range: 0.0867. Min Range: -0.0867

Max Act: 0.0540. Min Act: -0.0001

Data Index: 1306245 (C4 (Web Text))

Max Activating Token Index: 262