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

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

Neuron 991 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.8004. Min Act: -0.1700

Data Index: 1637239 (Python Code)

Max Activating Token Index: 664

Text #1

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.8317. Min Act: -0.1700

Data Index: 1212281 (C4 (Web Text))

Max Activating Token Index: 756

Text #2

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.7692. Min Act: -0.1700

Data Index: 1480306 (Python Code)

Max Activating Token Index: 743

Text #3

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.7379. Min Act: -0.1700

Data Index: 525785 (C4 (Web Text))

Max Activating Token Index: 883

Text #4

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.7379. Min Act: -0.1700

Data Index: 1625773 (Python Code)

Max Activating Token Index: 269

Text #5

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.7379. Min Act: -0.1700

Data Index: 1428106 (Python Code)

Max Activating Token Index: 98

Text #6

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.7379. Min Act: -0.1700

Data Index: 1619787 (Python Code)

Max Activating Token Index: 859

Text #7

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.7379. Min Act: -0.1700

Data Index: 1387311 (Python Code)

Max Activating Token Index: 545

Text #8

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.7066. Min Act: -0.1700

Data Index: 1551144 (Python Code)

Max Activating Token Index: 666

Text #9

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.7379. Min Act: -0.1700

Data Index: 641059 (C4 (Web Text))

Max Activating Token Index: 901

Text #10

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.7066. Min Act: -0.1700

Data Index: 1391042 (Python Code)

Max Activating Token Index: 145

Text #11

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.7066. Min Act: -0.1700

Data Index: 1614608 (Python Code)

Max Activating Token Index: 307

Text #12

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.7066. Min Act: -0.1700

Data Index: 442225 (C4 (Web Text))

Max Activating Token Index: 986

Text #13

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.6753. Min Act: -0.1700

Data Index: 1613008 (Python Code)

Max Activating Token Index: 63

Text #14

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.6440. Min Act: -0.1700

Data Index: 1646564 (Python Code)

Max Activating Token Index: 541

Text #15

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.6753. Min Act: -0.1700

Data Index: 1454607 (Python Code)

Max Activating Token Index: 134

Text #16

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.5970. Min Act: -0.1700

Data Index: 1509792 (Python Code)

Max Activating Token Index: 718

Text #17

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.6753. Min Act: -0.1700

Data Index: 1490910 (Python Code)

Max Activating Token Index: 76

Text #18

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.6440. Min Act: -0.1700

Data Index: 1540940 (Python Code)

Max Activating Token Index: 64

Text #19

Max Range: 3.8317. Min Range: -3.8317

Max Act: 3.6753. Min Act: -0.1700

Data Index: 1416062 (Python Code)

Max Activating Token Index: 140