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

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

Neuron 76 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: 5.5281. Min Range: -5.5281

Max Act: 5.5281. Min Act: -0.1700

Data Index: 638308 (C4 (Web Text))

Max Activating Token Index: 724

Text #1

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.3718. Min Act: -0.1700

Data Index: 1262514 (C4 (Web Text))

Max Activating Token Index: 893

Text #2

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.3093. Min Act: -0.1700

Data Index: 581107 (C4 (Web Text))

Max Activating Token Index: 359

Text #3

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.4031. Min Act: -0.1700

Data Index: 1346755 (C4 (Web Text))

Max Activating Token Index: 145

Text #4

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.3718. Min Act: -0.1700

Data Index: 915802 (C4 (Web Text))

Max Activating Token Index: 786

Text #5

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.3406. Min Act: -0.1700

Data Index: 857144 (C4 (Web Text))

Max Activating Token Index: 807

Text #6

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.2781. Min Act: -0.1700

Data Index: 707225 (C4 (Web Text))

Max Activating Token Index: 461

Text #7

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.3093. Min Act: -0.1700

Data Index: 225762 (C4 (Web Text))

Max Activating Token Index: 964

Text #8

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.2781. Min Act: -0.1700

Data Index: 1323890 (C4 (Web Text))

Max Activating Token Index: 693

Text #9

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.2156. Min Act: -0.1700

Data Index: 570104 (C4 (Web Text))

Max Activating Token Index: 884

Text #10

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.3093. Min Act: -0.1700

Data Index: 161331 (C4 (Web Text))

Max Activating Token Index: 69

Text #11

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.2781. Min Act: -0.1700

Data Index: 970018 (C4 (Web Text))

Max Activating Token Index: 590

Text #12

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.2468. Min Act: -0.1700

Data Index: 22664 (C4 (Web Text))

Max Activating Token Index: 438

Text #13

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.3093. Min Act: -0.1700

Data Index: 1246372 (C4 (Web Text))

Max Activating Token Index: 638

Text #14

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.2468. Min Act: -0.1700

Data Index: 269458 (C4 (Web Text))

Max Activating Token Index: 641

Text #15

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.2156. Min Act: -0.1700

Data Index: 924727 (C4 (Web Text))

Max Activating Token Index: 766

Text #16

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.2156. Min Act: -0.1700

Data Index: 685237 (C4 (Web Text))

Max Activating Token Index: 453

Text #17

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.1843. Min Act: -0.1700

Data Index: 1432315 (Python Code)

Max Activating Token Index: 709

Text #18

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.2156. Min Act: -0.1700

Data Index: 1135187 (C4 (Web Text))

Max Activating Token Index: 1003

Text #19

Max Range: 5.5281. Min Range: -5.5281

Max Act: 5.2156. Min Act: -0.1700

Data Index: 808988 (C4 (Web Text))

Max Activating Token Index: 431