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

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

Neuron 2950 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: 4.2667. Min Range: -4.2667

Max Act: 4.0219. Min Act: -0.0000

Data Index: 819275 (C4 (Web Text))

Max Activating Token Index: 625

Text #1

Max Range: 4.2667. Min Range: -4.2667

Max Act: 4.1127. Min Act: -0.0000

Data Index: 265691 (C4 (Web Text))

Max Activating Token Index: 849

Text #2

Max Range: 4.2667. Min Range: -4.2667

Max Act: 4.2667. Min Act: -0.0000

Data Index: 1232112 (C4 (Web Text))

Max Activating Token Index: 1001

Text #3

Max Range: 4.2667. Min Range: -4.2667

Max Act: 3.0648. Min Act: -0.0000

Data Index: 281948 (C4 (Web Text))

Max Activating Token Index: 293

Text #4

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.8227. Min Act: -0.0000

Data Index: 605274 (C4 (Web Text))

Max Activating Token Index: 975

Text #5

Max Range: 4.2667. Min Range: -4.2667

Max Act: 3.2747. Min Act: -0.0000

Data Index: 881915 (C4 (Web Text))

Max Activating Token Index: 280

Text #6

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.5179. Min Act: -0.0000

Data Index: 526789 (C4 (Web Text))

Max Activating Token Index: 54

Text #7

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.7052. Min Act: -0.0000

Data Index: 688496 (C4 (Web Text))

Max Activating Token Index: 507

Text #8

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.4933. Min Act: -0.0000

Data Index: 213084 (C4 (Web Text))

Max Activating Token Index: 724

Text #9

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.7107. Min Act: -0.0000

Data Index: 541907 (C4 (Web Text))

Max Activating Token Index: 715

Text #10

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.5650. Min Act: -0.0000

Data Index: 344495 (C4 (Web Text))

Max Activating Token Index: 264

Text #11

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.5124. Min Act: -0.0000

Data Index: 552204 (C4 (Web Text))

Max Activating Token Index: 891

Text #12

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.2084. Min Act: -0.0000

Data Index: 678377 (C4 (Web Text))

Max Activating Token Index: 989

Text #13

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.3140. Min Act: -0.0000

Data Index: 525901 (C4 (Web Text))

Max Activating Token Index: 902

Text #14

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.4240. Min Act: -0.0000

Data Index: 1107815 (C4 (Web Text))

Max Activating Token Index: 825

Text #15

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.3308. Min Act: -0.0000

Data Index: 499029 (C4 (Web Text))

Max Activating Token Index: 677

Text #16

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.3919. Min Act: -0.0000

Data Index: 748597 (C4 (Web Text))

Max Activating Token Index: 328

Text #17

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.1571. Min Act: -0.0000

Data Index: 781820 (C4 (Web Text))

Max Activating Token Index: 683

Text #18

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.0452. Min Act: -0.0000

Data Index: 389414 (C4 (Web Text))

Max Activating Token Index: 664

Text #19

Max Range: 4.2667. Min Range: -4.2667

Max Act: 2.3537. Min Act: -0.0000

Data Index: 1131523 (C4 (Web Text))

Max Activating Token Index: 535