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

Dataset: The Pile

Neuron 102 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 8.7803. Min Range: -8.7803

Max Act: 8.7803. Min Act: -0.0000

Data Index: 1589969 (The Pile)

Max Activating Token Index: 701

Text #1

Max Range: 8.7803. Min Range: -8.7803

Max Act: 7.1659. Min Act: -0.0000

Data Index: 696910 (The Pile)

Max Activating Token Index: 296

Text #2

Max Range: 8.7803. Min Range: -8.7803

Max Act: 6.7368. Min Act: -0.0000

Data Index: 116653 (The Pile)

Max Activating Token Index: 526

Text #3

Max Range: 8.7803. Min Range: -8.7803

Max Act: 6.2153. Min Act: -0.0000

Data Index: 943986 (The Pile)

Max Activating Token Index: 512

Text #4

Max Range: 8.7803. Min Range: -8.7803

Max Act: 5.1819. Min Act: -0.0000

Data Index: 1176578 (The Pile)

Max Activating Token Index: 813

Text #5

Max Range: 8.7803. Min Range: -8.7803

Max Act: 6.7588. Min Act: -0.0000

Data Index: 918941 (The Pile)

Max Activating Token Index: 243

Text #6

Max Range: 8.7803. Min Range: -8.7803

Max Act: 5.5737. Min Act: -0.0000

Data Index: 411240 (The Pile)

Max Activating Token Index: 636

Text #7

Max Range: 8.7803. Min Range: -8.7803

Max Act: 5.2034. Min Act: -0.0000

Data Index: 767802 (The Pile)

Max Activating Token Index: 875

Text #8

Max Range: 8.7803. Min Range: -8.7803

Max Act: 5.3182. Min Act: -0.0000

Data Index: 1447690 (The Pile)

Max Activating Token Index: 438

Text #9

Max Range: 8.7803. Min Range: -8.7803

Max Act: 5.4694. Min Act: -0.0000

Data Index: 204506 (The Pile)

Max Activating Token Index: 673

Text #10

Max Range: 8.7803. Min Range: -8.7803

Max Act: 5.0445. Min Act: -0.0000

Data Index: 775647 (The Pile)

Max Activating Token Index: 590

Text #11

Max Range: 8.7803. Min Range: -8.7803

Max Act: 5.0859. Min Act: -0.0000

Data Index: 304175 (The Pile)

Max Activating Token Index: 434

Text #12

Max Range: 8.7803. Min Range: -8.7803

Max Act: 5.2810. Min Act: -0.0000

Data Index: 259084 (The Pile)

Max Activating Token Index: 369

Text #13

Max Range: 8.7803. Min Range: -8.7803

Max Act: 5.4226. Min Act: -0.0000

Data Index: 1676530 (The Pile)

Max Activating Token Index: 839

Text #14

Max Range: 8.7803. Min Range: -8.7803

Max Act: 5.5839. Min Act: -0.0000

Data Index: 643878 (The Pile)

Max Activating Token Index: 720

Text #15

Max Range: 8.7803. Min Range: -8.7803

Max Act: 5.3777. Min Act: -0.0000

Data Index: 1516200 (The Pile)

Max Activating Token Index: 274

Text #16

Max Range: 8.7803. Min Range: -8.7803

Max Act: 4.8066. Min Act: -0.0000

Data Index: 965964 (The Pile)

Max Activating Token Index: 224

Text #17

Max Range: 8.7803. Min Range: -8.7803

Max Act: 5.1075. Min Act: -0.0000

Data Index: 657353 (The Pile)

Max Activating Token Index: 471

Text #18

Max Range: 8.7803. Min Range: -8.7803

Max Act: 4.3998. Min Act: -0.0000

Data Index: 966759 (The Pile)

Max Activating Token Index: 866

Text #19

Max Range: 8.7803. Min Range: -8.7803

Max Act: 4.4875. Min Act: -0.0000

Data Index: 1839135 (The Pile)

Max Activating Token Index: 472