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

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

Neuron 1996 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1730. Min Act: -0.1700

Data Index: 717801 (C4 (Web Text))

Max Activating Token Index: 321

Text #1

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1730. Min Act: -0.1700

Data Index: 215971 (C4 (Web Text))

Max Activating Token Index: 474

Text #2

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1730. Min Act: -0.1700

Data Index: 365087 (C4 (Web Text))

Max Activating Token Index: 528

Text #3

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1730. Min Act: -0.1700

Data Index: 1177064 (C4 (Web Text))

Max Activating Token Index: 235

Text #4

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1573. Min Act: -0.1700

Data Index: 822221 (C4 (Web Text))

Max Activating Token Index: 766

Text #5

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1573. Min Act: -0.1700

Data Index: 1057784 (C4 (Web Text))

Max Activating Token Index: 709

Text #6

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1573. Min Act: -0.1700

Data Index: 472009 (C4 (Web Text))

Max Activating Token Index: 140

Text #7

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1573. Min Act: -0.1700

Data Index: 184635 (C4 (Web Text))

Max Activating Token Index: 265

Text #8

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1415. Min Act: -0.1700

Data Index: 1261039 (C4 (Web Text))

Max Activating Token Index: 183

Text #9

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1258. Min Act: -0.1700

Data Index: 1540429 (Python Code)

Max Activating Token Index: 443

Text #10

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1100. Min Act: -0.1700

Data Index: 169310 (C4 (Web Text))

Max Activating Token Index: 842

Text #11

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1415. Min Act: -0.1700

Data Index: 280735 (C4 (Web Text))

Max Activating Token Index: 14

Text #12

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1258. Min Act: -0.1700

Data Index: 673622 (C4 (Web Text))

Max Activating Token Index: 861

Text #13

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1258. Min Act: -0.1700

Data Index: 726010 (C4 (Web Text))

Max Activating Token Index: 21

Text #14

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1100. Min Act: -0.1700

Data Index: 22044 (C4 (Web Text))

Max Activating Token Index: 179

Text #15

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1100. Min Act: -0.1700

Data Index: 409958 (C4 (Web Text))

Max Activating Token Index: 26

Text #16

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1100. Min Act: -0.1700

Data Index: 185277 (C4 (Web Text))

Max Activating Token Index: 795

Text #17

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1100. Min Act: -0.1700

Data Index: 6009 (C4 (Web Text))

Max Activating Token Index: 662

Text #18

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1100. Min Act: -0.1700

Data Index: 1450399 (Python Code)

Max Activating Token Index: 442

Text #19

Max Range: 3.1730. Min Range: -3.1730

Max Act: 3.1100. Min Act: -0.1700

Data Index: 1473526 (Python Code)

Max Activating Token Index: 911