Model: GELU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 565 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-3l')

Text #0

Max Range: 8.8642. Min Range: -8.8642

Max Act: 8.8642. Min Act: -0.1700

Data Index: 579710 (C4 (Web Text))

Max Activating Token Index: 158

Text #1

Max Range: 8.8642. Min Range: -8.8642

Max Act: 8.3642. Min Act: -0.1700

Data Index: 458674 (C4 (Web Text))

Max Activating Token Index: 468

Text #2

Max Range: 8.8642. Min Range: -8.8642

Max Act: 8.3642. Min Act: -0.1700

Data Index: 1447668 (Python Code)

Max Activating Token Index: 187

Text #3

Max Range: 8.8642. Min Range: -8.8642

Max Act: 8.2705. Min Act: -0.1700

Data Index: 1634098 (Python Code)

Max Activating Token Index: 1018

Text #4

Max Range: 8.8642. Min Range: -8.8642

Max Act: 8.1455. Min Act: -0.1700

Data Index: 1412600 (Python Code)

Max Activating Token Index: 922

Text #5

Max Range: 8.8642. Min Range: -8.8642

Max Act: 8.1142. Min Act: -0.1700

Data Index: 1381891 (Python Code)

Max Activating Token Index: 816

Text #6

Max Range: 8.8642. Min Range: -8.8642

Max Act: 8.0517. Min Act: -0.1700

Data Index: 1319415 (C4 (Web Text))

Max Activating Token Index: 260

Text #7

Max Range: 8.8642. Min Range: -8.8642

Max Act: 8.0205. Min Act: -0.1700

Data Index: 711940 (C4 (Web Text))

Max Activating Token Index: 205

Text #8

Max Range: 8.8642. Min Range: -8.8642

Max Act: 7.9267. Min Act: -0.1700

Data Index: 464246 (C4 (Web Text))

Max Activating Token Index: 465

Text #9

Max Range: 8.8642. Min Range: -8.8642

Max Act: 7.8955. Min Act: -0.1700

Data Index: 1543718 (Python Code)

Max Activating Token Index: 934

Text #10

Max Range: 8.8642. Min Range: -8.8642

Max Act: 7.9267. Min Act: -0.1700

Data Index: 1475142 (Python Code)

Max Activating Token Index: 216

Text #11

Max Range: 8.8642. Min Range: -8.8642

Max Act: 7.8017. Min Act: -0.1700

Data Index: 1647141 (Python Code)

Max Activating Token Index: 289

Text #12

Max Range: 8.8642. Min Range: -8.8642

Max Act: 7.8642. Min Act: -0.1700

Data Index: 1504645 (Python Code)

Max Activating Token Index: 189

Text #13

Max Range: 8.8642. Min Range: -8.8642

Max Act: 7.9267. Min Act: -0.1700

Data Index: 1532901 (Python Code)

Max Activating Token Index: 852

Text #14

Max Range: 8.8642. Min Range: -8.8642

Max Act: 7.8330. Min Act: -0.1700

Data Index: 1566010 (Python Code)

Max Activating Token Index: 807

Text #15

Max Range: 8.8642. Min Range: -8.8642

Max Act: 7.8642. Min Act: -0.1700

Data Index: 541527 (C4 (Web Text))

Max Activating Token Index: 550

Text #16

Max Range: 8.8642. Min Range: -8.8642

Max Act: 7.9267. Min Act: -0.1700

Data Index: 91610 (C4 (Web Text))

Max Activating Token Index: 737

Text #17

Max Range: 8.8642. Min Range: -8.8642

Max Act: 7.8642. Min Act: -0.1700

Data Index: 388219 (C4 (Web Text))

Max Activating Token Index: 118

Text #18

Max Range: 8.8642. Min Range: -8.8642

Max Act: 7.8955. Min Act: -0.1700

Data Index: 1634219 (Python Code)

Max Activating Token Index: 548

Text #19

Max Range: 8.8642. Min Range: -8.8642

Max Act: 7.8955. Min Act: -0.1700

Data Index: 1378535 (Python Code)

Max Activating Token Index: 106