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

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

Neuron 334 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: 5.6357. Min Range: -5.6357

Max Act: 5.6357. Min Act: -0.1696

Data Index: 1460292 (Python Code)

Max Activating Token Index: 156

Text #1

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.6357. Min Act: -0.1698

Data Index: 1577723 (Python Code)

Max Activating Token Index: 776

Text #2

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.5732. Min Act: -0.1699

Data Index: 1569924 (Python Code)

Max Activating Token Index: 531

Text #3

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.5420. Min Act: -0.1699

Data Index: 1390364 (Python Code)

Max Activating Token Index: 336

Text #4

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.5107. Min Act: -0.1699

Data Index: 1453204 (Python Code)

Max Activating Token Index: 240

Text #5

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.5107. Min Act: -0.1698

Data Index: 1638377 (Python Code)

Max Activating Token Index: 659

Text #6

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.4795. Min Act: -0.1686

Data Index: 1496581 (Python Code)

Max Activating Token Index: 894

Text #7

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.5107. Min Act: -0.1236

Data Index: 358423 (C4 (Web Text))

Max Activating Token Index: 870

Text #8

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.4170. Min Act: -0.1700

Data Index: 1504040 (Python Code)

Max Activating Token Index: 796

Text #9

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.4170. Min Act: -0.1699

Data Index: 1445785 (Python Code)

Max Activating Token Index: 854

Text #10

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.4170. Min Act: -0.1699

Data Index: 1498918 (Python Code)

Max Activating Token Index: 175

Text #11

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.3857. Min Act: -0.1699

Data Index: 1401573 (Python Code)

Max Activating Token Index: 339

Text #12

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.3545. Min Act: -0.1699

Data Index: 1476988 (Python Code)

Max Activating Token Index: 471

Text #13

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.3857. Min Act: -0.1700

Data Index: 1527950 (Python Code)

Max Activating Token Index: 904

Text #14

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.3545. Min Act: -0.1569

Data Index: 1482189 (Python Code)

Max Activating Token Index: 67

Text #15

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.3545. Min Act: -0.1699

Data Index: 1619755 (Python Code)

Max Activating Token Index: 1010

Text #16

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.3545. Min Act: -0.1699

Data Index: 1654957 (Python Code)

Max Activating Token Index: 14

Text #17

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.3232. Min Act: -0.1699

Data Index: 1645072 (Python Code)

Max Activating Token Index: 961

Text #18

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.3232. Min Act: -0.1700

Data Index: 1451802 (Python Code)

Max Activating Token Index: 686

Text #19

Max Range: 5.6357. Min Range: -5.6357

Max Act: 5.3232. Min Act: -0.1700

Data Index: 273220 (C4 (Web Text))

Max Activating Token Index: 521