Model: GELU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 1050 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-4l')

Text #0

Max Range: 7.4926. Min Range: -7.4926

Max Act: 7.4926. Min Act: -0.1694

Data Index: 260751 (C4 (Web Text))

Max Activating Token Index: 479

Text #1

Max Range: 7.4926. Min Range: -7.4926

Max Act: 7.4301. Min Act: -0.1699

Data Index: 1362821 (Python Code)

Max Activating Token Index: 601

Text #2

Max Range: 7.4926. Min Range: -7.4926

Max Act: 7.3676. Min Act: -0.1699

Data Index: 461443 (C4 (Web Text))

Max Activating Token Index: 606

Text #3

Max Range: 7.4926. Min Range: -7.4926

Max Act: 7.2426. Min Act: -0.1696

Data Index: 973677 (C4 (Web Text))

Max Activating Token Index: 680

Text #4

Max Range: 7.4926. Min Range: -7.4926

Max Act: 7.0239. Min Act: -0.1699

Data Index: 39595 (C4 (Web Text))

Max Activating Token Index: 268

Text #5

Max Range: 7.4926. Min Range: -7.4926

Max Act: 7.1176. Min Act: -0.1700

Data Index: 1529017 (Python Code)

Max Activating Token Index: 977

Text #6

Max Range: 7.4926. Min Range: -7.4926

Max Act: 7.0551. Min Act: -0.1700

Data Index: 1571066 (Python Code)

Max Activating Token Index: 635

Text #7

Max Range: 7.4926. Min Range: -7.4926

Max Act: 6.9926. Min Act: -0.1699

Data Index: 943458 (C4 (Web Text))

Max Activating Token Index: 764

Text #8

Max Range: 7.4926. Min Range: -7.4926

Max Act: 6.9614. Min Act: -0.1699

Data Index: 1602086 (Python Code)

Max Activating Token Index: 407

Text #9

Max Range: 7.4926. Min Range: -7.4926

Max Act: 6.9301. Min Act: -0.1688

Data Index: 1509506 (Python Code)

Max Activating Token Index: 96

Text #10

Max Range: 7.4926. Min Range: -7.4926

Max Act: 6.9614. Min Act: -0.1697

Data Index: 573041 (C4 (Web Text))

Max Activating Token Index: 36

Text #11

Max Range: 7.4926. Min Range: -7.4926

Max Act: 6.8989. Min Act: -0.1698

Data Index: 1643446 (Python Code)

Max Activating Token Index: 248

Text #12

Max Range: 7.4926. Min Range: -7.4926

Max Act: 6.8989. Min Act: -0.1699

Data Index: 1641013 (Python Code)

Max Activating Token Index: 459

Text #13

Max Range: 7.4926. Min Range: -7.4926

Max Act: 6.8989. Min Act: -0.1628

Data Index: 1454783 (Python Code)

Max Activating Token Index: 517

Text #14

Max Range: 7.4926. Min Range: -7.4926

Max Act: 6.8364. Min Act: -0.1699

Data Index: 228471 (C4 (Web Text))

Max Activating Token Index: 262

Text #15

Max Range: 7.4926. Min Range: -7.4926

Max Act: 6.8676. Min Act: -0.1700

Data Index: 1066775 (C4 (Web Text))

Max Activating Token Index: 837

Text #16

Max Range: 7.4926. Min Range: -7.4926

Max Act: 6.8364. Min Act: -0.1675

Data Index: 977376 (C4 (Web Text))

Max Activating Token Index: 692

Text #17

Max Range: 7.4926. Min Range: -7.4926

Max Act: 6.8364. Min Act: -0.1699

Data Index: 1500823 (Python Code)

Max Activating Token Index: 244

Text #18

Max Range: 7.4926. Min Range: -7.4926

Max Act: 6.8051. Min Act: -0.1700

Data Index: 1433828 (Python Code)

Max Activating Token Index: 330

Text #19

Max Range: 7.4926. Min Range: -7.4926

Max Act: 6.8364. Min Act: -0.1700

Data Index: 542708 (C4 (Web Text))

Max Activating Token Index: 512