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

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

Neuron 618 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: 8.3429. Min Range: -8.3429

Max Act: 8.2804. Min Act: -0.1700

Data Index: 451416 (C4 (Web Text))

Max Activating Token Index: 735

Text #1

Max Range: 8.3429. Min Range: -8.3429

Max Act: 8.3429. Min Act: -0.1700

Data Index: 1077559 (C4 (Web Text))

Max Activating Token Index: 894

Text #2

Max Range: 8.3429. Min Range: -8.3429

Max Act: 8.3429. Min Act: -0.1700

Data Index: 1233276 (C4 (Web Text))

Max Activating Token Index: 613

Text #3

Max Range: 8.3429. Min Range: -8.3429

Max Act: 8.0929. Min Act: -0.1700

Data Index: 1578990 (Python Code)

Max Activating Token Index: 742

Text #4

Max Range: 8.3429. Min Range: -8.3429

Max Act: 8.1554. Min Act: -0.1700

Data Index: 1387777 (Python Code)

Max Activating Token Index: 305

Text #5

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.9679. Min Act: -0.1699

Data Index: 153971 (C4 (Web Text))

Max Activating Token Index: 788

Text #6

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.9679. Min Act: -0.1700

Data Index: 822854 (C4 (Web Text))

Max Activating Token Index: 735

Text #7

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.9054. Min Act: -0.1700

Data Index: 1561318 (Python Code)

Max Activating Token Index: 324

Text #8

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.9679. Min Act: -0.1700

Data Index: 554543 (C4 (Web Text))

Max Activating Token Index: 915

Text #9

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.9679. Min Act: -0.1700

Data Index: 916731 (C4 (Web Text))

Max Activating Token Index: 922

Text #10

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.9679. Min Act: -0.1700

Data Index: 356376 (C4 (Web Text))

Max Activating Token Index: 895

Text #11

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.9054. Min Act: -0.1700

Data Index: 789373 (C4 (Web Text))

Max Activating Token Index: 883

Text #12

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.8429. Min Act: -0.1700

Data Index: 523317 (C4 (Web Text))

Max Activating Token Index: 212

Text #13

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.8429. Min Act: -0.1700

Data Index: 1375698 (Python Code)

Max Activating Token Index: 394

Text #14

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.8429. Min Act: -0.1700

Data Index: 1563801 (Python Code)

Max Activating Token Index: 704

Text #15

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.9054. Min Act: -0.1700

Data Index: 941933 (C4 (Web Text))

Max Activating Token Index: 571

Text #16

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.7804. Min Act: -0.1700

Data Index: 92377 (C4 (Web Text))

Max Activating Token Index: 680

Text #17

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.7804. Min Act: -0.1700

Data Index: 891114 (C4 (Web Text))

Max Activating Token Index: 982

Text #18

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.7804. Min Act: -0.1699

Data Index: 982942 (C4 (Web Text))

Max Activating Token Index: 893

Text #19

Max Range: 8.3429. Min Range: -8.3429

Max Act: 7.7804. Min Act: -0.1700

Data Index: 1483142 (Python Code)

Max Activating Token Index: 764