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

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

Neuron 344 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: 3.7478. Min Range: -3.7478

Max Act: 3.7478. Min Act: -0.1700

Data Index: 1380509 (Python Code)

Max Activating Token Index: 1023

Text #1

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.7165. Min Act: -0.1700

Data Index: 1597181 (Python Code)

Max Activating Token Index: 1023

Text #2

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.5756. Min Act: -0.1700

Data Index: 1071793 (C4 (Web Text))

Max Activating Token Index: 908

Text #3

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.5130. Min Act: -0.1700

Data Index: 1253676 (C4 (Web Text))

Max Activating Token Index: 132

Text #4

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.4816. Min Act: -0.1700

Data Index: 1539798 (Python Code)

Max Activating Token Index: 275

Text #5

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.4816. Min Act: -0.1700

Data Index: 44313 (C4 (Web Text))

Max Activating Token Index: 955

Text #6

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.4973. Min Act: -0.1700

Data Index: 217154 (C4 (Web Text))

Max Activating Token Index: 276

Text #7

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.4659. Min Act: -0.1700

Data Index: 1620117 (Python Code)

Max Activating Token Index: 72

Text #8

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.4346. Min Act: -0.1700

Data Index: 139762 (C4 (Web Text))

Max Activating Token Index: 968

Text #9

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.3719. Min Act: -0.1700

Data Index: 1641466 (Python Code)

Max Activating Token Index: 466

Text #10

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.4032. Min Act: -0.1700

Data Index: 962749 (C4 (Web Text))

Max Activating Token Index: 1023

Text #11

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.4189. Min Act: -0.1700

Data Index: 1453120 (Python Code)

Max Activating Token Index: 406

Text #12

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.4346. Min Act: -0.1700

Data Index: 1369096 (Python Code)

Max Activating Token Index: 707

Text #13

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.4032. Min Act: -0.1700

Data Index: 1512785 (Python Code)

Max Activating Token Index: 480

Text #14

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.4032. Min Act: -0.1700

Data Index: 1419221 (Python Code)

Max Activating Token Index: 426

Text #15

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.4032. Min Act: -0.1700

Data Index: 1637782 (Python Code)

Max Activating Token Index: 364

Text #16

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.4189. Min Act: -0.1700

Data Index: 1543573 (Python Code)

Max Activating Token Index: 456

Text #17

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.3875. Min Act: -0.1700

Data Index: 1451900 (Python Code)

Max Activating Token Index: 833

Text #18

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.3719. Min Act: -0.1700

Data Index: 1479290 (Python Code)

Max Activating Token Index: 411

Text #19

Max Range: 3.7478. Min Range: -3.7478

Max Act: 3.3562. Min Act: -0.1700

Data Index: 1504231 (Python Code)

Max Activating Token Index: 874