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

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

Neuron 2563 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.5666. Min Act: -0.0000

Data Index: 1069343 (C4 (Web Text))

Max Activating Token Index: 706

Text #1

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.5217. Min Act: -0.0000

Data Index: 969966 (C4 (Web Text))

Max Activating Token Index: 797

Text #2

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.5444. Min Act: -0.0000

Data Index: 1050380 (C4 (Web Text))

Max Activating Token Index: 841

Text #3

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.5159. Min Act: -0.0000

Data Index: 1122037 (C4 (Web Text))

Max Activating Token Index: 318

Text #4

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.5450. Min Act: -0.0000

Data Index: 956134 (C4 (Web Text))

Max Activating Token Index: 235

Text #5

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.4958. Min Act: -0.0000

Data Index: 1037135 (C4 (Web Text))

Max Activating Token Index: 1001

Text #6

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.4879. Min Act: -0.0000

Data Index: 750383 (C4 (Web Text))

Max Activating Token Index: 193

Text #7

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.5244. Min Act: -0.0000

Data Index: 409038 (C4 (Web Text))

Max Activating Token Index: 897

Text #8

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.4991. Min Act: -0.0000

Data Index: 15211 (C4 (Web Text))

Max Activating Token Index: 915

Text #9

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.5117. Min Act: -0.0000

Data Index: 1273232 (C4 (Web Text))

Max Activating Token Index: 507

Text #10

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.4925. Min Act: -0.0000

Data Index: 772246 (C4 (Web Text))

Max Activating Token Index: 303

Text #11

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.4881. Min Act: -0.0000

Data Index: 491342 (C4 (Web Text))

Max Activating Token Index: 665

Text #12

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.4514. Min Act: -0.0000

Data Index: 605280 (C4 (Web Text))

Max Activating Token Index: 798

Text #13

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.4951. Min Act: -0.0001

Data Index: 186216 (C4 (Web Text))

Max Activating Token Index: 863

Text #14

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.4623. Min Act: -0.0000

Data Index: 1583041 (Python Code)

Max Activating Token Index: 910

Text #15

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.4544. Min Act: -0.0000

Data Index: 1012416 (C4 (Web Text))

Max Activating Token Index: 115

Text #16

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.4842. Min Act: -0.0000

Data Index: 377478 (C4 (Web Text))

Max Activating Token Index: 858

Text #17

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.4496. Min Act: -0.0000

Data Index: 1271448 (C4 (Web Text))

Max Activating Token Index: 853

Text #18

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.4670. Min Act: -0.0000

Data Index: 666475 (C4 (Web Text))

Max Activating Token Index: 799

Text #19

Max Range: 0.5666. Min Range: -0.5666

Max Act: 0.4293. Min Act: -0.0000

Data Index: 617474 (C4 (Web Text))

Max Activating Token Index: 234