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

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

Neuron 1354 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-4l')

Text #0

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.5565. Min Act: -0.1700

Data Index: 1016379 (C4 (Web Text))

Max Activating Token Index: 890

Text #1

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.4625. Min Act: -0.1700

Data Index: 164562 (C4 (Web Text))

Max Activating Token Index: 482

Text #2

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.4625. Min Act: -0.1700

Data Index: 1077938 (C4 (Web Text))

Max Activating Token Index: 174

Text #3

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.4939. Min Act: -0.1700

Data Index: 475631 (C4 (Web Text))

Max Activating Token Index: 672

Text #4

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.3998. Min Act: -0.1700

Data Index: 1322517 (C4 (Web Text))

Max Activating Token Index: 729

Text #5

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.4312. Min Act: -0.1700

Data Index: 1403637 (Python Code)

Max Activating Token Index: 321

Text #6

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.3684. Min Act: -0.1700

Data Index: 1652748 (Python Code)

Max Activating Token Index: 685

Text #7

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.3370. Min Act: -0.1700

Data Index: 156842 (C4 (Web Text))

Max Activating Token Index: 881

Text #8

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.4312. Min Act: -0.1700

Data Index: 649052 (C4 (Web Text))

Max Activating Token Index: 165

Text #9

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.3684. Min Act: -0.1700

Data Index: 1048075 (C4 (Web Text))

Max Activating Token Index: 964

Text #10

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.3684. Min Act: -0.1700

Data Index: 555883 (C4 (Web Text))

Max Activating Token Index: 149

Text #11

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.3684. Min Act: -0.1700

Data Index: 311041 (C4 (Web Text))

Max Activating Token Index: 485

Text #12

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.3684. Min Act: -0.1700

Data Index: 1509465 (Python Code)

Max Activating Token Index: 597

Text #13

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.3370. Min Act: -0.1700

Data Index: 964995 (C4 (Web Text))

Max Activating Token Index: 182

Text #14

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.3213. Min Act: -0.1700

Data Index: 1656607 (Python Code)

Max Activating Token Index: 113

Text #15

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.2899. Min Act: -0.1700

Data Index: 88835 (C4 (Web Text))

Max Activating Token Index: 851

Text #16

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.3213. Min Act: -0.1700

Data Index: 795332 (C4 (Web Text))

Max Activating Token Index: 973

Text #17

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.3056. Min Act: -0.1700

Data Index: 169772 (C4 (Web Text))

Max Activating Token Index: 626

Text #18

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.3370. Min Act: -0.1700

Data Index: 493726 (C4 (Web Text))

Max Activating Token Index: 901

Text #19

Max Range: 3.5565. Min Range: -3.5565

Max Act: 3.3213. Min Act: -0.1700

Data Index: 421514 (C4 (Web Text))

Max Activating Token Index: 214