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

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

Neuron 1628 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.6231. Min Range: -5.6231

Max Act: 5.6231. Min Act: -0.1700

Data Index: 92139 (C4 (Web Text))

Max Activating Token Index: 25

Text #1

Max Range: 5.6231. Min Range: -5.6231

Max Act: 5.4669. Min Act: -0.1700

Data Index: 38020 (C4 (Web Text))

Max Activating Token Index: 101

Text #2

Max Range: 5.6231. Min Range: -5.6231

Max Act: 5.3731. Min Act: -0.1700

Data Index: 917841 (C4 (Web Text))

Max Activating Token Index: 813

Text #3

Max Range: 5.6231. Min Range: -5.6231

Max Act: 5.2169. Min Act: -0.1700

Data Index: 216731 (C4 (Web Text))

Max Activating Token Index: 123

Text #4

Max Range: 5.6231. Min Range: -5.6231

Max Act: 5.2481. Min Act: -0.1700

Data Index: 854827 (C4 (Web Text))

Max Activating Token Index: 1013

Text #5

Max Range: 5.6231. Min Range: -5.6231

Max Act: 5.1856. Min Act: -0.1700

Data Index: 1233227 (C4 (Web Text))

Max Activating Token Index: 321

Text #6

Max Range: 5.6231. Min Range: -5.6231

Max Act: 5.0606. Min Act: -0.1700

Data Index: 595499 (C4 (Web Text))

Max Activating Token Index: 891

Text #7

Max Range: 5.6231. Min Range: -5.6231

Max Act: 5.0919. Min Act: -0.1700

Data Index: 1067227 (C4 (Web Text))

Max Activating Token Index: 328

Text #8

Max Range: 5.6231. Min Range: -5.6231

Max Act: 5.0919. Min Act: -0.1700

Data Index: 1026085 (C4 (Web Text))

Max Activating Token Index: 475

Text #9

Max Range: 5.6231. Min Range: -5.6231

Max Act: 5.0919. Min Act: -0.1700

Data Index: 1033618 (C4 (Web Text))

Max Activating Token Index: 58

Text #10

Max Range: 5.6231. Min Range: -5.6231

Max Act: 5.0294. Min Act: -0.1700

Data Index: 1132881 (C4 (Web Text))

Max Activating Token Index: 896

Text #11

Max Range: 5.6231. Min Range: -5.6231

Max Act: 5.1231. Min Act: -0.1700

Data Index: 775971 (C4 (Web Text))

Max Activating Token Index: 731

Text #12

Max Range: 5.6231. Min Range: -5.6231

Max Act: 5.0294. Min Act: -0.1700

Data Index: 31478 (C4 (Web Text))

Max Activating Token Index: 598

Text #13

Max Range: 5.6231. Min Range: -5.6231

Max Act: 4.9981. Min Act: -0.1700

Data Index: 1026529 (C4 (Web Text))

Max Activating Token Index: 826

Text #14

Max Range: 5.6231. Min Range: -5.6231

Max Act: 5.0294. Min Act: -0.1700

Data Index: 48535 (C4 (Web Text))

Max Activating Token Index: 249

Text #15

Max Range: 5.6231. Min Range: -5.6231

Max Act: 4.9669. Min Act: -0.1700

Data Index: 981946 (C4 (Web Text))

Max Activating Token Index: 980

Text #16

Max Range: 5.6231. Min Range: -5.6231

Max Act: 4.9981. Min Act: -0.1700

Data Index: 164625 (C4 (Web Text))

Max Activating Token Index: 899

Text #17

Max Range: 5.6231. Min Range: -5.6231

Max Act: 4.9356. Min Act: -0.1700

Data Index: 745096 (C4 (Web Text))

Max Activating Token Index: 627

Text #18

Max Range: 5.6231. Min Range: -5.6231

Max Act: 4.9669. Min Act: -0.1700

Data Index: 747476 (C4 (Web Text))

Max Activating Token Index: 443

Text #19

Max Range: 5.6231. Min Range: -5.6231

Max Act: 4.9356. Min Act: -0.1700

Data Index: 474965 (C4 (Web Text))

Max Activating Token Index: 653