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

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

Neuron 1376 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: 5.1125. Min Range: -5.1125

Max Act: 5.1125. Min Act: -0.1700

Data Index: 75224 (C4 (Web Text))

Max Activating Token Index: 942

Text #1

Max Range: 5.1125. Min Range: -5.1125

Max Act: 5.0188. Min Act: -0.1700

Data Index: 118676 (C4 (Web Text))

Max Activating Token Index: 922

Text #2

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.9250. Min Act: -0.1700

Data Index: 873506 (C4 (Web Text))

Max Activating Token Index: 771

Text #3

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.7688. Min Act: -0.1700

Data Index: 97655 (C4 (Web Text))

Max Activating Token Index: 65

Text #4

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.7063. Min Act: -0.1700

Data Index: 707633 (C4 (Web Text))

Max Activating Token Index: 891

Text #5

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.7375. Min Act: -0.1700

Data Index: 630869 (C4 (Web Text))

Max Activating Token Index: 437

Text #6

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.6438. Min Act: -0.1700

Data Index: 955598 (C4 (Web Text))

Max Activating Token Index: 848

Text #7

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.6125. Min Act: -0.1700

Data Index: 439667 (C4 (Web Text))

Max Activating Token Index: 111

Text #8

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.5813. Min Act: -0.1700

Data Index: 65205 (C4 (Web Text))

Max Activating Token Index: 113

Text #9

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.5813. Min Act: -0.1700

Data Index: 69977 (C4 (Web Text))

Max Activating Token Index: 16

Text #10

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.5500. Min Act: -0.1700

Data Index: 1041083 (C4 (Web Text))

Max Activating Token Index: 118

Text #11

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.5187. Min Act: -0.1700

Data Index: 566535 (C4 (Web Text))

Max Activating Token Index: 25

Text #12

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.5500. Min Act: -0.1700

Data Index: 1110474 (C4 (Web Text))

Max Activating Token Index: 16

Text #13

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.4562. Min Act: -0.1700

Data Index: 538370 (C4 (Web Text))

Max Activating Token Index: 870

Text #14

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.4875. Min Act: -0.1700

Data Index: 1009622 (C4 (Web Text))

Max Activating Token Index: 385

Text #15

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.5187. Min Act: -0.1700

Data Index: 916028 (C4 (Web Text))

Max Activating Token Index: 55

Text #16

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.5187. Min Act: -0.1700

Data Index: 170304 (C4 (Web Text))

Max Activating Token Index: 73

Text #17

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.5187. Min Act: -0.1700

Data Index: 752914 (C4 (Web Text))

Max Activating Token Index: 230

Text #18

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.4875. Min Act: -0.1700

Data Index: 1084950 (C4 (Web Text))

Max Activating Token Index: 173

Text #19

Max Range: 5.1125. Min Range: -5.1125

Max Act: 4.4250. Min Act: -0.1700

Data Index: 1649537 (Python Code)

Max Activating Token Index: 1015