Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 2020 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 6.1686. Min Range: -6.1686

Max Act: 6.1686. Min Act: -0.0001

Data Index: 1492689 (Python Code)

Max Activating Token Index: 925

Text #1

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.8230. Min Act: -0.0001

Data Index: 1279716 (C4 (Web Text))

Max Activating Token Index: 1011

Text #2

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.7572. Min Act: -0.0001

Data Index: 607284 (C4 (Web Text))

Max Activating Token Index: 919

Text #3

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.7275. Min Act: -0.0001

Data Index: 752357 (C4 (Web Text))

Max Activating Token Index: 84

Text #4

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.4711. Min Act: -0.0001

Data Index: 1035894 (C4 (Web Text))

Max Activating Token Index: 837

Text #5

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.4126. Min Act: -0.0001

Data Index: 1604737 (Python Code)

Max Activating Token Index: 415

Text #6

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.5521. Min Act: -0.0001

Data Index: 607558 (C4 (Web Text))

Max Activating Token Index: 443

Text #7

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.5495. Min Act: -0.0001

Data Index: 1560621 (Python Code)

Max Activating Token Index: 983

Text #8

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.5704. Min Act: -0.0001

Data Index: 1566806 (Python Code)

Max Activating Token Index: 514

Text #9

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.7354. Min Act: -0.0001

Data Index: 655116 (C4 (Web Text))

Max Activating Token Index: 601

Text #10

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.1249. Min Act: -0.0001

Data Index: 347851 (C4 (Web Text))

Max Activating Token Index: 473

Text #11

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.3678. Min Act: -0.0001

Data Index: 1390199 (Python Code)

Max Activating Token Index: 830

Text #12

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.3555. Min Act: -0.0001

Data Index: 1409567 (Python Code)

Max Activating Token Index: 458

Text #13

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.1502. Min Act: -0.0001

Data Index: 1526650 (Python Code)

Max Activating Token Index: 792

Text #14

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.1086. Min Act: -0.0001

Data Index: 1027071 (C4 (Web Text))

Max Activating Token Index: 493

Text #15

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.3785. Min Act: -0.0001

Data Index: 1254562 (C4 (Web Text))

Max Activating Token Index: 619

Text #16

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.1721. Min Act: -0.0001

Data Index: 1485471 (Python Code)

Max Activating Token Index: 726

Text #17

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.2137. Min Act: -0.0001

Data Index: 448748 (C4 (Web Text))

Max Activating Token Index: 803

Text #18

Max Range: 6.1686. Min Range: -6.1686

Max Act: 4.9495. Min Act: -0.0001

Data Index: 739573 (C4 (Web Text))

Max Activating Token Index: 243

Text #19

Max Range: 6.1686. Min Range: -6.1686

Max Act: 5.3529. Min Act: -0.0001

Data Index: 333580 (C4 (Web Text))

Max Activating Token Index: 772