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

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

Neuron 453 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: 7.8025. Min Range: -7.8025

Max Act: 7.8025. Min Act: -0.1700

Data Index: 1532878 (Python Code)

Max Activating Token Index: 871

Text #1

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.5525. Min Act: -0.1700

Data Index: 1532499 (Python Code)

Max Activating Token Index: 312

Text #2

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.3650. Min Act: -0.1700

Data Index: 1510234 (Python Code)

Max Activating Token Index: 987

Text #3

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.5212. Min Act: -0.1700

Data Index: 1553902 (Python Code)

Max Activating Token Index: 576

Text #4

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.4275. Min Act: -0.1700

Data Index: 1473097 (Python Code)

Max Activating Token Index: 708

Text #5

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.3962. Min Act: -0.1700

Data Index: 1422446 (Python Code)

Max Activating Token Index: 557

Text #6

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.4275. Min Act: -0.1700

Data Index: 1573600 (Python Code)

Max Activating Token Index: 160

Text #7

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.4275. Min Act: -0.1700

Data Index: 1525786 (Python Code)

Max Activating Token Index: 383

Text #8

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.3962. Min Act: -0.1700

Data Index: 1379723 (Python Code)

Max Activating Token Index: 561

Text #9

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.3337. Min Act: -0.1700

Data Index: 1437723 (Python Code)

Max Activating Token Index: 287

Text #10

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.3962. Min Act: -0.1700

Data Index: 1645700 (Python Code)

Max Activating Token Index: 377

Text #11

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.3337. Min Act: -0.1700

Data Index: 1647853 (Python Code)

Max Activating Token Index: 619

Text #12

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.3962. Min Act: -0.1700

Data Index: 1471772 (Python Code)

Max Activating Token Index: 158

Text #13

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.3337. Min Act: -0.1700

Data Index: 1435642 (Python Code)

Max Activating Token Index: 248

Text #14

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.2712. Min Act: -0.1700

Data Index: 1390369 (Python Code)

Max Activating Token Index: 497

Text #15

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.2400. Min Act: -0.1700

Data Index: 1624961 (Python Code)

Max Activating Token Index: 441

Text #16

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.2400. Min Act: -0.1700

Data Index: 1584520 (Python Code)

Max Activating Token Index: 127

Text #17

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.3337. Min Act: -0.1700

Data Index: 1416626 (Python Code)

Max Activating Token Index: 160

Text #18

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.2087. Min Act: -0.1700

Data Index: 1418012 (Python Code)

Max Activating Token Index: 698

Text #19

Max Range: 7.8025. Min Range: -7.8025

Max Act: 7.2400. Min Act: -0.1700

Data Index: 1628510 (Python Code)

Max Activating Token Index: 140