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

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

Neuron 275 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: 5.1007. Min Range: -5.1007

Max Act: 5.1007. Min Act: -0.1700

Data Index: 215219 (C4 (Web Text))

Max Activating Token Index: 91

Text #1

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.7570. Min Act: -0.1700

Data Index: 479021 (C4 (Web Text))

Max Activating Token Index: 144

Text #2

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.5695. Min Act: -0.1699

Data Index: 571944 (C4 (Web Text))

Max Activating Token Index: 660

Text #3

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.6007. Min Act: -0.1700

Data Index: 489433 (C4 (Web Text))

Max Activating Token Index: 18

Text #4

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.5695. Min Act: -0.1700

Data Index: 865419 (C4 (Web Text))

Max Activating Token Index: 79

Text #5

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.5382. Min Act: -0.1700

Data Index: 1087936 (C4 (Web Text))

Max Activating Token Index: 39

Text #6

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.5070. Min Act: -0.1700

Data Index: 346021 (C4 (Web Text))

Max Activating Token Index: 473

Text #7

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.5070. Min Act: -0.1700

Data Index: 207343 (C4 (Web Text))

Max Activating Token Index: 58

Text #8

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.4445. Min Act: -0.1700

Data Index: 527997 (C4 (Web Text))

Max Activating Token Index: 130

Text #9

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.4445. Min Act: -0.1700

Data Index: 267374 (C4 (Web Text))

Max Activating Token Index: 500

Text #10

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.4445. Min Act: -0.1700

Data Index: 924510 (C4 (Web Text))

Max Activating Token Index: 388

Text #11

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.4445. Min Act: -0.1700

Data Index: 1089237 (C4 (Web Text))

Max Activating Token Index: 312

Text #12

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.4445. Min Act: -0.1700

Data Index: 5765 (C4 (Web Text))

Max Activating Token Index: 9

Text #13

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.4132. Min Act: -0.1700

Data Index: 822968 (C4 (Web Text))

Max Activating Token Index: 698

Text #14

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.4132. Min Act: -0.1700

Data Index: 886196 (C4 (Web Text))

Max Activating Token Index: 36

Text #15

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.4132. Min Act: -0.1700

Data Index: 916105 (C4 (Web Text))

Max Activating Token Index: 378

Text #16

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.3820. Min Act: -0.1700

Data Index: 232319 (C4 (Web Text))

Max Activating Token Index: 693

Text #17

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.4132. Min Act: -0.1700

Data Index: 246925 (C4 (Web Text))

Max Activating Token Index: 916

Text #18

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.4132. Min Act: -0.1700

Data Index: 1036214 (C4 (Web Text))

Max Activating Token Index: 496

Text #19

Max Range: 5.1007. Min Range: -5.1007

Max Act: 4.4445. Min Act: -0.1700

Data Index: 702537 (C4 (Web Text))

Max Activating Token Index: 192