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

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

Neuron 1245 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: 3.5214. Min Range: -3.5214

Max Act: 3.5214. Min Act: -0.1700

Data Index: 1653219 (Python Code)

Max Activating Token Index: 945

Text #1

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.5057. Min Act: -0.1700

Data Index: 1619149 (Python Code)

Max Activating Token Index: 975

Text #2

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.5057. Min Act: -0.1700

Data Index: 152226 (C4 (Web Text))

Max Activating Token Index: 549

Text #3

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.5214. Min Act: -0.1700

Data Index: 1619561 (Python Code)

Max Activating Token Index: 945

Text #4

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.5057. Min Act: -0.1700

Data Index: 358601 (C4 (Web Text))

Max Activating Token Index: 826

Text #5

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.5214. Min Act: -0.1700

Data Index: 397672 (C4 (Web Text))

Max Activating Token Index: 797

Text #6

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.4744. Min Act: -0.1700

Data Index: 1389048 (Python Code)

Max Activating Token Index: 849

Text #7

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.5057. Min Act: -0.1700

Data Index: 1651635 (Python Code)

Max Activating Token Index: 906

Text #8

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.4900. Min Act: -0.1700

Data Index: 1364956 (Python Code)

Max Activating Token Index: 702

Text #9

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.4900. Min Act: -0.1700

Data Index: 1629864 (Python Code)

Max Activating Token Index: 653

Text #10

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.4587. Min Act: -0.1700

Data Index: 1560843 (Python Code)

Max Activating Token Index: 814

Text #11

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.4587. Min Act: -0.1700

Data Index: 1384313 (Python Code)

Max Activating Token Index: 650

Text #12

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.5214. Min Act: -0.1700

Data Index: 1357410 (C4 (Web Text))

Max Activating Token Index: 555

Text #13

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.4587. Min Act: -0.1700

Data Index: 1410649 (Python Code)

Max Activating Token Index: 109

Text #14

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.4744. Min Act: -0.1700

Data Index: 1648926 (Python Code)

Max Activating Token Index: 665

Text #15

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.5057. Min Act: -0.1700

Data Index: 1619008 (Python Code)

Max Activating Token Index: 583

Text #16

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.4587. Min Act: -0.1700

Data Index: 1608600 (Python Code)

Max Activating Token Index: 328

Text #17

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.4744. Min Act: -0.1700

Data Index: 1427319 (Python Code)

Max Activating Token Index: 750

Text #18

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.4587. Min Act: -0.1700

Data Index: 1379652 (Python Code)

Max Activating Token Index: 848

Text #19

Max Range: 3.5214. Min Range: -3.5214

Max Act: 3.4744. Min Act: -0.1700

Data Index: 1574046 (Python Code)

Max Activating Token Index: 910