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

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

Neuron 1148 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: 5.4231. Min Range: -5.4231

Max Act: 5.4231. Min Act: -0.1700

Data Index: 917365 (C4 (Web Text))

Max Activating Token Index: 309

Text #1

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.2981. Min Act: -0.1700

Data Index: 139129 (C4 (Web Text))

Max Activating Token Index: 637

Text #2

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.2668. Min Act: -0.1700

Data Index: 287247 (C4 (Web Text))

Max Activating Token Index: 683

Text #3

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.1731. Min Act: -0.1700

Data Index: 1087864 (C4 (Web Text))

Max Activating Token Index: 529

Text #4

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.2043. Min Act: -0.1700

Data Index: 737495 (C4 (Web Text))

Max Activating Token Index: 554

Text #5

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.1418. Min Act: -0.1700

Data Index: 1342958 (C4 (Web Text))

Max Activating Token Index: 772

Text #6

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.1106. Min Act: -0.1700

Data Index: 79703 (C4 (Web Text))

Max Activating Token Index: 696

Text #7

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.1106. Min Act: -0.1700

Data Index: 1066451 (C4 (Web Text))

Max Activating Token Index: 172

Text #8

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.1106. Min Act: -0.1700

Data Index: 756664 (C4 (Web Text))

Max Activating Token Index: 718

Text #9

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.0793. Min Act: -0.1700

Data Index: 702807 (C4 (Web Text))

Max Activating Token Index: 336

Text #10

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.0793. Min Act: -0.1700

Data Index: 882235 (C4 (Web Text))

Max Activating Token Index: 967

Text #11

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.0481. Min Act: -0.1700

Data Index: 369327 (C4 (Web Text))

Max Activating Token Index: 71

Text #12

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.0793. Min Act: -0.1700

Data Index: 98415 (C4 (Web Text))

Max Activating Token Index: 709

Text #13

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.0168. Min Act: -0.1700

Data Index: 883344 (C4 (Web Text))

Max Activating Token Index: 910

Text #14

Max Range: 5.4231. Min Range: -5.4231

Max Act: 4.9856. Min Act: -0.1700

Data Index: 852790 (C4 (Web Text))

Max Activating Token Index: 644

Text #15

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.0481. Min Act: -0.1700

Data Index: 1056176 (C4 (Web Text))

Max Activating Token Index: 816

Text #16

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.0168. Min Act: -0.1700

Data Index: 515430 (C4 (Web Text))

Max Activating Token Index: 226

Text #17

Max Range: 5.4231. Min Range: -5.4231

Max Act: 4.9856. Min Act: -0.1700

Data Index: 523666 (C4 (Web Text))

Max Activating Token Index: 47

Text #18

Max Range: 5.4231. Min Range: -5.4231

Max Act: 5.0481. Min Act: -0.1700

Data Index: 490196 (C4 (Web Text))

Max Activating Token Index: 384

Text #19

Max Range: 5.4231. Min Range: -5.4231

Max Act: 4.9543. Min Act: -0.1700

Data Index: 29312 (C4 (Web Text))

Max Activating Token Index: 329