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

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

Neuron 856 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: 9.5986. Min Range: -9.5986

Max Act: 9.5986. Min Act: -0.1700

Data Index: 463671 (C4 (Web Text))

Max Activating Token Index: 645

Text #1

Max Range: 9.5986. Min Range: -9.5986

Max Act: 9.5361. Min Act: -0.1700

Data Index: 1155231 (C4 (Web Text))

Max Activating Token Index: 802

Text #2

Max Range: 9.5986. Min Range: -9.5986

Max Act: 9.0361. Min Act: -0.1700

Data Index: 300016 (C4 (Web Text))

Max Activating Token Index: 815

Text #3

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.7236. Min Act: -0.1700

Data Index: 659292 (C4 (Web Text))

Max Activating Token Index: 884

Text #4

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.4111. Min Act: -0.1700

Data Index: 1096528 (C4 (Web Text))

Max Activating Token Index: 945

Text #5

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.5986. Min Act: -0.1700

Data Index: 13774 (C4 (Web Text))

Max Activating Token Index: 368

Text #6

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.5361. Min Act: -0.1700

Data Index: 961482 (C4 (Web Text))

Max Activating Token Index: 270

Text #7

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.1923. Min Act: -0.1700

Data Index: 536988 (C4 (Web Text))

Max Activating Token Index: 773

Text #8

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.4736. Min Act: -0.1700

Data Index: 834351 (C4 (Web Text))

Max Activating Token Index: 711

Text #9

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.5986. Min Act: -0.1700

Data Index: 1219707 (C4 (Web Text))

Max Activating Token Index: 791

Text #10

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.3798. Min Act: -0.1700

Data Index: 217852 (C4 (Web Text))

Max Activating Token Index: 710

Text #11

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.2861. Min Act: -0.1700

Data Index: 45358 (C4 (Web Text))

Max Activating Token Index: 1006

Text #12

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.1611. Min Act: -0.1700

Data Index: 813562 (C4 (Web Text))

Max Activating Token Index: 698

Text #13

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.3798. Min Act: -0.1700

Data Index: 1016269 (C4 (Web Text))

Max Activating Token Index: 535

Text #14

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.3798. Min Act: -0.1700

Data Index: 1000551 (C4 (Web Text))

Max Activating Token Index: 600

Text #15

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.2236. Min Act: -0.1700

Data Index: 900973 (C4 (Web Text))

Max Activating Token Index: 763

Text #16

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.2861. Min Act: -0.1700

Data Index: 383563 (C4 (Web Text))

Max Activating Token Index: 634

Text #17

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.1298. Min Act: -0.1700

Data Index: 1167425 (C4 (Web Text))

Max Activating Token Index: 29

Text #18

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.0986. Min Act: -0.1700

Data Index: 1100507 (C4 (Web Text))

Max Activating Token Index: 65

Text #19

Max Range: 9.5986. Min Range: -9.5986

Max Act: 8.3173. Min Act: -0.1700

Data Index: 88956 (C4 (Web Text))

Max Activating Token Index: 909