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

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

Neuron 1995 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 8.6237. Min Range: -8.6237

Max Act: 8.6237. Min Act: -0.1700

Data Index: 531482 (C4 (Web Text))

Max Activating Token Index: 97

Text #1

Max Range: 8.6237. Min Range: -8.6237

Max Act: 8.3737. Min Act: -0.1700

Data Index: 116907 (C4 (Web Text))

Max Activating Token Index: 540

Text #2

Max Range: 8.6237. Min Range: -8.6237

Max Act: 8.0925. Min Act: -0.1700

Data Index: 364307 (C4 (Web Text))

Max Activating Token Index: 507

Text #3

Max Range: 8.6237. Min Range: -8.6237

Max Act: 8.0925. Min Act: -0.1700

Data Index: 643059 (C4 (Web Text))

Max Activating Token Index: 417

Text #4

Max Range: 8.6237. Min Range: -8.6237

Max Act: 8.0612. Min Act: -0.1700

Data Index: 643313 (C4 (Web Text))

Max Activating Token Index: 249

Text #5

Max Range: 8.6237. Min Range: -8.6237

Max Act: 8.0300. Min Act: -0.1700

Data Index: 1121590 (C4 (Web Text))

Max Activating Token Index: 303

Text #6

Max Range: 8.6237. Min Range: -8.6237

Max Act: 7.9987. Min Act: -0.1700

Data Index: 619882 (C4 (Web Text))

Max Activating Token Index: 409

Text #7

Max Range: 8.6237. Min Range: -8.6237

Max Act: 7.8737. Min Act: -0.1700

Data Index: 777794 (C4 (Web Text))

Max Activating Token Index: 987

Text #8

Max Range: 8.6237. Min Range: -8.6237

Max Act: 7.9050. Min Act: -0.1700

Data Index: 102345 (C4 (Web Text))

Max Activating Token Index: 905

Text #9

Max Range: 8.6237. Min Range: -8.6237

Max Act: 7.9362. Min Act: -0.1700

Data Index: 626208 (C4 (Web Text))

Max Activating Token Index: 815

Text #10

Max Range: 8.6237. Min Range: -8.6237

Max Act: 7.9362. Min Act: -0.1700

Data Index: 234889 (C4 (Web Text))

Max Activating Token Index: 1000

Text #11

Max Range: 8.6237. Min Range: -8.6237

Max Act: 7.9362. Min Act: -0.1700

Data Index: 1459073 (Python Code)

Max Activating Token Index: 704

Text #12

Max Range: 8.6237. Min Range: -8.6237

Max Act: 7.8425. Min Act: -0.1700

Data Index: 1308052 (C4 (Web Text))

Max Activating Token Index: 596

Text #13

Max Range: 8.6237. Min Range: -8.6237

Max Act: 7.9050. Min Act: -0.1700

Data Index: 493359 (C4 (Web Text))

Max Activating Token Index: 910

Text #14

Max Range: 8.6237. Min Range: -8.6237

Max Act: 7.8112. Min Act: -0.1700

Data Index: 692100 (C4 (Web Text))

Max Activating Token Index: 389

Text #15

Max Range: 8.6237. Min Range: -8.6237

Max Act: 7.8737. Min Act: -0.1700

Data Index: 945659 (C4 (Web Text))

Max Activating Token Index: 84

Text #16

Max Range: 8.6237. Min Range: -8.6237

Max Act: 7.8425. Min Act: -0.1700

Data Index: 749017 (C4 (Web Text))

Max Activating Token Index: 297

Text #17

Max Range: 8.6237. Min Range: -8.6237

Max Act: 7.8425. Min Act: -0.1700

Data Index: 1236512 (C4 (Web Text))

Max Activating Token Index: 410

Text #18

Max Range: 8.6237. Min Range: -8.6237

Max Act: 7.7800. Min Act: -0.1700

Data Index: 843498 (C4 (Web Text))

Max Activating Token Index: 404

Text #19

Max Range: 8.6237. Min Range: -8.6237

Max Act: 7.7487. Min Act: -0.1700

Data Index: 721585 (C4 (Web Text))

Max Activating Token Index: 108