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

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

Neuron 1333 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.9326. Min Act: -0.1700

Data Index: 1404748 (Python Code)

Max Activating Token Index: 552

Text #1

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.4951. Min Act: -0.1700

Data Index: 1595908 (Python Code)

Max Activating Token Index: 95

Text #2

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.4326. Min Act: -0.1700

Data Index: 1585900 (Python Code)

Max Activating Token Index: 707

Text #3

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.3701. Min Act: -0.1700

Data Index: 1390839 (Python Code)

Max Activating Token Index: 43

Text #4

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.1201. Min Act: -0.1700

Data Index: 95724 (C4 (Web Text))

Max Activating Token Index: 274

Text #5

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.0888. Min Act: -0.1699

Data Index: 410526 (C4 (Web Text))

Max Activating Token Index: 621

Text #6

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.1201. Min Act: -0.1700

Data Index: 1557301 (Python Code)

Max Activating Token Index: 384

Text #7

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.1826. Min Act: -0.1700

Data Index: 433008 (C4 (Web Text))

Max Activating Token Index: 531

Text #8

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.0576. Min Act: -0.1700

Data Index: 1562109 (Python Code)

Max Activating Token Index: 992

Text #9

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.0263. Min Act: -0.1700

Data Index: 1565623 (Python Code)

Max Activating Token Index: 859

Text #10

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.0888. Min Act: -0.1700

Data Index: 852831 (C4 (Web Text))

Max Activating Token Index: 803

Text #11

Max Range: 7.9326. Min Range: -7.9326

Max Act: 6.9951. Min Act: -0.1700

Data Index: 242658 (C4 (Web Text))

Max Activating Token Index: 706

Text #12

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.1201. Min Act: -0.1700

Data Index: 232011 (C4 (Web Text))

Max Activating Token Index: 966

Text #13

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.1201. Min Act: -0.1700

Data Index: 1476856 (Python Code)

Max Activating Token Index: 121

Text #14

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.1201. Min Act: -0.1700

Data Index: 1049116 (C4 (Web Text))

Max Activating Token Index: 832

Text #15

Max Range: 7.9326. Min Range: -7.9326

Max Act: 6.9951. Min Act: -0.1700

Data Index: 1462900 (Python Code)

Max Activating Token Index: 27

Text #16

Max Range: 7.9326. Min Range: -7.9326

Max Act: 6.9951. Min Act: -0.1700

Data Index: 518877 (C4 (Web Text))

Max Activating Token Index: 225

Text #17

Max Range: 7.9326. Min Range: -7.9326

Max Act: 7.0263. Min Act: -0.1700

Data Index: 1126588 (C4 (Web Text))

Max Activating Token Index: 329

Text #18

Max Range: 7.9326. Min Range: -7.9326

Max Act: 6.9951. Min Act: -0.1699

Data Index: 1285930 (C4 (Web Text))

Max Activating Token Index: 676

Text #19

Max Range: 7.9326. Min Range: -7.9326

Max Act: 6.9638. Min Act: -0.1700

Data Index: 1654650 (Python Code)

Max Activating Token Index: 290