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

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

Neuron 1803 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.3915. Min Range: -3.3915

Max Act: 3.3915. Min Act: -0.1700

Data Index: 555426 (C4 (Web Text))

Max Activating Token Index: 644

Text #1

Max Range: 3.3915. Min Range: -3.3915

Max Act: 3.2502. Min Act: -0.1700

Data Index: 1078129 (C4 (Web Text))

Max Activating Token Index: 825

Text #2

Max Range: 3.3915. Min Range: -3.3915

Max Act: 3.0928. Min Act: -0.1700

Data Index: 634663 (C4 (Web Text))

Max Activating Token Index: 495

Text #3

Max Range: 3.3915. Min Range: -3.3915

Max Act: 3.0770. Min Act: -0.1700

Data Index: 532428 (C4 (Web Text))

Max Activating Token Index: 118

Text #4

Max Range: 3.3915. Min Range: -3.3915

Max Act: 3.0928. Min Act: -0.1700

Data Index: 1045553 (C4 (Web Text))

Max Activating Token Index: 984

Text #5

Max Range: 3.3915. Min Range: -3.3915

Max Act: 3.0296. Min Act: -0.1700

Data Index: 1297473 (C4 (Web Text))

Max Activating Token Index: 476

Text #6

Max Range: 3.3915. Min Range: -3.3915

Max Act: 3.0296. Min Act: -0.1700

Data Index: 58771 (C4 (Web Text))

Max Activating Token Index: 579

Text #7

Max Range: 3.3915. Min Range: -3.3915

Max Act: 3.0454. Min Act: -0.1700

Data Index: 1292889 (C4 (Web Text))

Max Activating Token Index: 875

Text #8

Max Range: 3.3915. Min Range: -3.3915

Max Act: 3.0296. Min Act: -0.1700

Data Index: 1262929 (C4 (Web Text))

Max Activating Token Index: 379

Text #9

Max Range: 3.3915. Min Range: -3.3915

Max Act: 2.9822. Min Act: -0.1700

Data Index: 732289 (C4 (Web Text))

Max Activating Token Index: 382

Text #10

Max Range: 3.3915. Min Range: -3.3915

Max Act: 3.0138. Min Act: -0.1700

Data Index: 1154431 (C4 (Web Text))

Max Activating Token Index: 490

Text #11

Max Range: 3.3915. Min Range: -3.3915

Max Act: 2.9664. Min Act: -0.1700

Data Index: 198488 (C4 (Web Text))

Max Activating Token Index: 324

Text #12

Max Range: 3.3915. Min Range: -3.3915

Max Act: 2.9189. Min Act: -0.1700

Data Index: 1026134 (C4 (Web Text))

Max Activating Token Index: 858

Text #13

Max Range: 3.3915. Min Range: -3.3915

Max Act: 2.9506. Min Act: -0.1700

Data Index: 921978 (C4 (Web Text))

Max Activating Token Index: 121

Text #14

Max Range: 3.3915. Min Range: -3.3915

Max Act: 2.9347. Min Act: -0.1700

Data Index: 638851 (C4 (Web Text))

Max Activating Token Index: 484

Text #15

Max Range: 3.3915. Min Range: -3.3915

Max Act: 2.9347. Min Act: -0.1700

Data Index: 198071 (C4 (Web Text))

Max Activating Token Index: 131

Text #16

Max Range: 3.3915. Min Range: -3.3915

Max Act: 2.8872. Min Act: -0.1700

Data Index: 440528 (C4 (Web Text))

Max Activating Token Index: 221

Text #17

Max Range: 3.3915. Min Range: -3.3915

Max Act: 2.9030. Min Act: -0.1700

Data Index: 1537444 (Python Code)

Max Activating Token Index: 104

Text #18

Max Range: 3.3915. Min Range: -3.3915

Max Act: 2.9030. Min Act: -0.1700

Data Index: 344790 (C4 (Web Text))

Max Activating Token Index: 672

Text #19

Max Range: 3.3915. Min Range: -3.3915

Max Act: 2.8395. Min Act: -0.1700

Data Index: 357653 (C4 (Web Text))

Max Activating Token Index: 725