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

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

Neuron 1271 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: 8.5665. Min Range: -8.5665

Max Act: 8.5665. Min Act: -0.1700

Data Index: 215579 (C4 (Web Text))

Max Activating Token Index: 210

Text #1

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.5665. Min Act: -0.1700

Data Index: 652598 (C4 (Web Text))

Max Activating Token Index: 635

Text #2

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.5040. Min Act: -0.1700

Data Index: 1606221 (Python Code)

Max Activating Token Index: 118

Text #3

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.3477. Min Act: -0.1700

Data Index: 1314972 (C4 (Web Text))

Max Activating Token Index: 785

Text #4

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.2540. Min Act: -0.1700

Data Index: 1017495 (C4 (Web Text))

Max Activating Token Index: 886

Text #5

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.2227. Min Act: -0.1700

Data Index: 875627 (C4 (Web Text))

Max Activating Token Index: 258

Text #6

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.1915. Min Act: -0.1700

Data Index: 613472 (C4 (Web Text))

Max Activating Token Index: 972

Text #7

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.1602. Min Act: -0.1700

Data Index: 958785 (C4 (Web Text))

Max Activating Token Index: 1021

Text #8

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.2227. Min Act: -0.1700

Data Index: 41700 (C4 (Web Text))

Max Activating Token Index: 645

Text #9

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.1915. Min Act: -0.1700

Data Index: 535640 (C4 (Web Text))

Max Activating Token Index: 163

Text #10

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.1915. Min Act: -0.1700

Data Index: 645467 (C4 (Web Text))

Max Activating Token Index: 198

Text #11

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.1290. Min Act: -0.1700

Data Index: 752749 (C4 (Web Text))

Max Activating Token Index: 895

Text #12

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.1290. Min Act: -0.1700

Data Index: 91209 (C4 (Web Text))

Max Activating Token Index: 727

Text #13

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.1290. Min Act: -0.1700

Data Index: 493227 (C4 (Web Text))

Max Activating Token Index: 950

Text #14

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.0977. Min Act: -0.1700

Data Index: 1090263 (C4 (Web Text))

Max Activating Token Index: 123

Text #15

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.0665. Min Act: -0.1700

Data Index: 1460763 (Python Code)

Max Activating Token Index: 764

Text #16

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.1290. Min Act: -0.1700

Data Index: 580854 (C4 (Web Text))

Max Activating Token Index: 1001

Text #17

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.0977. Min Act: -0.1700

Data Index: 1167756 (C4 (Web Text))

Max Activating Token Index: 516

Text #18

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.0040. Min Act: -0.1700

Data Index: 261364 (C4 (Web Text))

Max Activating Token Index: 862

Text #19

Max Range: 8.5665. Min Range: -8.5665

Max Act: 8.0352. Min Act: -0.1700

Data Index: 211498 (C4 (Web Text))

Max Activating Token Index: 172