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

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

Neuron 768 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.5829. Min Act: -0.0000

Data Index: 1246489 (C4 (Web Text))

Max Activating Token Index: 388

Text #1

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.5623. Min Act: -0.0000

Data Index: 439763 (C4 (Web Text))

Max Activating Token Index: 724

Text #2

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.4996. Min Act: -0.0000

Data Index: 177426 (C4 (Web Text))

Max Activating Token Index: 819

Text #3

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.5085. Min Act: -0.0000

Data Index: 898415 (C4 (Web Text))

Max Activating Token Index: 867

Text #4

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.4849. Min Act: -0.0000

Data Index: 214454 (C4 (Web Text))

Max Activating Token Index: 214

Text #5

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.4643. Min Act: -0.0000

Data Index: 332948 (C4 (Web Text))

Max Activating Token Index: 129

Text #6

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.4365. Min Act: -0.0000

Data Index: 794010 (C4 (Web Text))

Max Activating Token Index: 235

Text #7

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.4326. Min Act: -0.0000

Data Index: 56498 (C4 (Web Text))

Max Activating Token Index: 918

Text #8

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.4301. Min Act: -0.0000

Data Index: 1245121 (C4 (Web Text))

Max Activating Token Index: 820

Text #9

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.4252. Min Act: -0.0000

Data Index: 1140493 (C4 (Web Text))

Max Activating Token Index: 612

Text #10

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.4073. Min Act: -0.0000

Data Index: 1399340 (Python Code)

Max Activating Token Index: 315

Text #11

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.4393. Min Act: -0.0000

Data Index: 1120958 (C4 (Web Text))

Max Activating Token Index: 250

Text #12

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.3700. Min Act: -0.0000

Data Index: 640750 (C4 (Web Text))

Max Activating Token Index: 310

Text #13

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.4120. Min Act: -0.0000

Data Index: 1162657 (C4 (Web Text))

Max Activating Token Index: 337

Text #14

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.3574. Min Act: -0.0000

Data Index: 455762 (C4 (Web Text))

Max Activating Token Index: 644

Text #15

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.4049. Min Act: -0.0000

Data Index: 1257704 (C4 (Web Text))

Max Activating Token Index: 81

Text #16

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.3818. Min Act: -0.0000

Data Index: 1335940 (C4 (Web Text))

Max Activating Token Index: 57

Text #17

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.3689. Min Act: -0.0000

Data Index: 112404 (C4 (Web Text))

Max Activating Token Index: 523

Text #18

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.3923. Min Act: -0.0000

Data Index: 609756 (C4 (Web Text))

Max Activating Token Index: 250

Text #19

Max Range: 0.5829. Min Range: -0.5829

Max Act: 0.4279. Min Act: -0.0000

Data Index: 527213 (C4 (Web Text))

Max Activating Token Index: 803