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

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

Neuron 929 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.1782. Min Range: -3.1782

Max Act: 3.1782. Min Act: -0.1700

Data Index: 1537234 (Python Code)

Max Activating Token Index: 63

Text #1

Max Range: 3.1782. Min Range: -3.1782

Max Act: 3.1782. Min Act: -0.1700

Data Index: 666265 (C4 (Web Text))

Max Activating Token Index: 19

Text #2

Max Range: 3.1782. Min Range: -3.1782

Max Act: 3.1625. Min Act: -0.1700

Data Index: 351512 (C4 (Web Text))

Max Activating Token Index: 836

Text #3

Max Range: 3.1782. Min Range: -3.1782

Max Act: 3.0521. Min Act: -0.1700

Data Index: 574027 (C4 (Web Text))

Max Activating Token Index: 289

Text #4

Max Range: 3.1782. Min Range: -3.1782

Max Act: 3.0679. Min Act: -0.1700

Data Index: 82387 (C4 (Web Text))

Max Activating Token Index: 547

Text #5

Max Range: 3.1782. Min Range: -3.1782

Max Act: 3.0363. Min Act: -0.1700

Data Index: 1620084 (Python Code)

Max Activating Token Index: 226

Text #6

Max Range: 3.1782. Min Range: -3.1782

Max Act: 2.9889. Min Act: -0.1700

Data Index: 91222 (C4 (Web Text))

Max Activating Token Index: 983

Text #7

Max Range: 3.1782. Min Range: -3.1782

Max Act: 3.0047. Min Act: -0.1700

Data Index: 263088 (C4 (Web Text))

Max Activating Token Index: 790

Text #8

Max Range: 3.1782. Min Range: -3.1782

Max Act: 2.9573. Min Act: -0.1700

Data Index: 648147 (C4 (Web Text))

Max Activating Token Index: 428

Text #9

Max Range: 3.1782. Min Range: -3.1782

Max Act: 2.9731. Min Act: -0.1700

Data Index: 325653 (C4 (Web Text))

Max Activating Token Index: 975

Text #10

Max Range: 3.1782. Min Range: -3.1782

Max Act: 2.9731. Min Act: -0.1700

Data Index: 556145 (C4 (Web Text))

Max Activating Token Index: 821

Text #11

Max Range: 3.1782. Min Range: -3.1782

Max Act: 2.9731. Min Act: -0.1700

Data Index: 932009 (C4 (Web Text))

Max Activating Token Index: 1012

Text #12

Max Range: 3.1782. Min Range: -3.1782

Max Act: 2.9731. Min Act: -0.1700

Data Index: 1479324 (Python Code)

Max Activating Token Index: 265

Text #13

Max Range: 3.1782. Min Range: -3.1782

Max Act: 2.9414. Min Act: -0.1700

Data Index: 1008119 (C4 (Web Text))

Max Activating Token Index: 533

Text #14

Max Range: 3.1782. Min Range: -3.1782

Max Act: 2.9256. Min Act: -0.1700

Data Index: 231441 (C4 (Web Text))

Max Activating Token Index: 34

Text #15

Max Range: 3.1782. Min Range: -3.1782

Max Act: 2.8303. Min Act: -0.1700

Data Index: 1157037 (C4 (Web Text))

Max Activating Token Index: 607

Text #16

Max Range: 3.1782. Min Range: -3.1782

Max Act: 2.8144. Min Act: -0.1700

Data Index: 376356 (C4 (Web Text))

Max Activating Token Index: 745

Text #17

Max Range: 3.1782. Min Range: -3.1782

Max Act: 2.8303. Min Act: -0.1700

Data Index: 351711 (C4 (Web Text))

Max Activating Token Index: 710

Text #18

Max Range: 3.1782. Min Range: -3.1782

Max Act: 2.8462. Min Act: -0.1700

Data Index: 769272 (C4 (Web Text))

Max Activating Token Index: 513

Text #19

Max Range: 3.1782. Min Range: -3.1782

Max Act: 2.8462. Min Act: -0.1700

Data Index: 1634891 (Python Code)

Max Activating Token Index: 31