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

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

Neuron 481 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: 8.7230. Min Range: -8.7230

Max Act: 8.7230. Min Act: -0.1700

Data Index: 161255 (C4 (Web Text))

Max Activating Token Index: 65

Text #1

Max Range: 8.7230. Min Range: -8.7230

Max Act: 8.2855. Min Act: -0.1700

Data Index: 1590105 (Python Code)

Max Activating Token Index: 296

Text #2

Max Range: 8.7230. Min Range: -8.7230

Max Act: 8.3480. Min Act: -0.1700

Data Index: 1546897 (Python Code)

Max Activating Token Index: 505

Text #3

Max Range: 8.7230. Min Range: -8.7230

Max Act: 8.2230. Min Act: -0.1700

Data Index: 1455531 (Python Code)

Max Activating Token Index: 592

Text #4

Max Range: 8.7230. Min Range: -8.7230

Max Act: 8.0980. Min Act: -0.1700

Data Index: 1361000 (Python Code)

Max Activating Token Index: 906

Text #5

Max Range: 8.7230. Min Range: -8.7230

Max Act: 8.1605. Min Act: -0.1700

Data Index: 460347 (C4 (Web Text))

Max Activating Token Index: 747

Text #6

Max Range: 8.7230. Min Range: -8.7230

Max Act: 8.0668. Min Act: -0.1700

Data Index: 1613292 (Python Code)

Max Activating Token Index: 507

Text #7

Max Range: 8.7230. Min Range: -8.7230

Max Act: 8.0355. Min Act: -0.1700

Data Index: 1440706 (Python Code)

Max Activating Token Index: 312

Text #8

Max Range: 8.7230. Min Range: -8.7230

Max Act: 8.0355. Min Act: -0.1700

Data Index: 1586114 (Python Code)

Max Activating Token Index: 714

Text #9

Max Range: 8.7230. Min Range: -8.7230

Max Act: 8.0355. Min Act: -0.1700

Data Index: 1490425 (Python Code)

Max Activating Token Index: 559

Text #10

Max Range: 8.7230. Min Range: -8.7230

Max Act: 8.0355. Min Act: -0.1700

Data Index: 1452844 (Python Code)

Max Activating Token Index: 480

Text #11

Max Range: 8.7230. Min Range: -8.7230

Max Act: 7.9418. Min Act: -0.1700

Data Index: 1517397 (Python Code)

Max Activating Token Index: 576

Text #12

Max Range: 8.7230. Min Range: -8.7230

Max Act: 7.9105. Min Act: -0.1700

Data Index: 1546885 (Python Code)

Max Activating Token Index: 787

Text #13

Max Range: 8.7230. Min Range: -8.7230

Max Act: 7.9105. Min Act: -0.1700

Data Index: 1520699 (Python Code)

Max Activating Token Index: 320

Text #14

Max Range: 8.7230. Min Range: -8.7230

Max Act: 7.9418. Min Act: -0.1700

Data Index: 1438420 (Python Code)

Max Activating Token Index: 892

Text #15

Max Range: 8.7230. Min Range: -8.7230

Max Act: 7.9105. Min Act: -0.1700

Data Index: 1635321 (Python Code)

Max Activating Token Index: 300

Text #16

Max Range: 8.7230. Min Range: -8.7230

Max Act: 7.8480. Min Act: -0.1700

Data Index: 1402941 (Python Code)

Max Activating Token Index: 656

Text #17

Max Range: 8.7230. Min Range: -8.7230

Max Act: 7.7543. Min Act: -0.1699

Data Index: 1226311 (C4 (Web Text))

Max Activating Token Index: 141

Text #18

Max Range: 8.7230. Min Range: -8.7230

Max Act: 7.8793. Min Act: -0.1700

Data Index: 1636270 (Python Code)

Max Activating Token Index: 725

Text #19

Max Range: 8.7230. Min Range: -8.7230

Max Act: 7.7543. Min Act: -0.1699

Data Index: 1653635 (Python Code)

Max Activating Token Index: 158