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

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

Neuron 3072 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: 4.0080. Min Range: -4.0080

Max Act: 4.0080. Min Act: -0.0000

Data Index: 48510 (C4 (Web Text))

Max Activating Token Index: 384

Text #1

Max Range: 4.0080. Min Range: -4.0080

Max Act: 3.4596. Min Act: -0.0000

Data Index: 297625 (C4 (Web Text))

Max Activating Token Index: 124

Text #2

Max Range: 4.0080. Min Range: -4.0080

Max Act: 3.3767. Min Act: -0.0000

Data Index: 896690 (C4 (Web Text))

Max Activating Token Index: 930

Text #3

Max Range: 4.0080. Min Range: -4.0080

Max Act: 3.5169. Min Act: -0.0000

Data Index: 995386 (C4 (Web Text))

Max Activating Token Index: 318

Text #4

Max Range: 4.0080. Min Range: -4.0080

Max Act: 2.9078. Min Act: -0.0000

Data Index: 657170 (C4 (Web Text))

Max Activating Token Index: 442

Text #5

Max Range: 4.0080. Min Range: -4.0080

Max Act: 2.7995. Min Act: -0.0000

Data Index: 1005392 (C4 (Web Text))

Max Activating Token Index: 620

Text #6

Max Range: 4.0080. Min Range: -4.0080

Max Act: 2.7620. Min Act: -0.0000

Data Index: 531825 (C4 (Web Text))

Max Activating Token Index: 841

Text #7

Max Range: 4.0080. Min Range: -4.0080

Max Act: 2.6743. Min Act: -0.0000

Data Index: 1189935 (C4 (Web Text))

Max Activating Token Index: 496

Text #8

Max Range: 4.0080. Min Range: -4.0080

Max Act: 2.5009. Min Act: -0.0000

Data Index: 673789 (C4 (Web Text))

Max Activating Token Index: 201

Text #9

Max Range: 4.0080. Min Range: -4.0080

Max Act: 2.5829. Min Act: -0.0000

Data Index: 842939 (C4 (Web Text))

Max Activating Token Index: 473

Text #10

Max Range: 4.0080. Min Range: -4.0080

Max Act: 2.4960. Min Act: -0.0000

Data Index: 33412 (C4 (Web Text))

Max Activating Token Index: 797

Text #11

Max Range: 4.0080. Min Range: -4.0080

Max Act: 2.3671. Min Act: -0.0000

Data Index: 457583 (C4 (Web Text))

Max Activating Token Index: 533

Text #12

Max Range: 4.0080. Min Range: -4.0080

Max Act: 2.1666. Min Act: -0.0000

Data Index: 269284 (C4 (Web Text))

Max Activating Token Index: 59

Text #13

Max Range: 4.0080. Min Range: -4.0080

Max Act: 1.9861. Min Act: -0.0000

Data Index: 475261 (C4 (Web Text))

Max Activating Token Index: 710

Text #14

Max Range: 4.0080. Min Range: -4.0080

Max Act: 1.9181. Min Act: -0.0000

Data Index: 415402 (C4 (Web Text))

Max Activating Token Index: 74

Text #15

Max Range: 4.0080. Min Range: -4.0080

Max Act: 1.9011. Min Act: -0.0000

Data Index: 684106 (C4 (Web Text))

Max Activating Token Index: 466

Text #16

Max Range: 4.0080. Min Range: -4.0080

Max Act: 1.8339. Min Act: -0.0000

Data Index: 620806 (C4 (Web Text))

Max Activating Token Index: 444

Text #17

Max Range: 4.0080. Min Range: -4.0080

Max Act: 1.8260. Min Act: -0.0000

Data Index: 103083 (C4 (Web Text))

Max Activating Token Index: 25

Text #18

Max Range: 4.0080. Min Range: -4.0080

Max Act: 1.7583. Min Act: -0.0000

Data Index: 527114 (C4 (Web Text))

Max Activating Token Index: 594

Text #19

Max Range: 4.0080. Min Range: -4.0080

Max Act: 1.6893. Min Act: -0.0000

Data Index: 95738 (C4 (Web Text))

Max Activating Token Index: 363