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

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

Neuron 661 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-3l')

Text #0

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.9071. Min Act: -0.1700

Data Index: 501481 (C4 (Web Text))

Max Activating Token Index: 327

Text #1

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.8445. Min Act: -0.1700

Data Index: 973302 (C4 (Web Text))

Max Activating Token Index: 721

Text #2

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.7820. Min Act: -0.1700

Data Index: 688171 (C4 (Web Text))

Max Activating Token Index: 769

Text #3

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.7507. Min Act: -0.1700

Data Index: 926684 (C4 (Web Text))

Max Activating Token Index: 415

Text #4

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.7194. Min Act: -0.1700

Data Index: 170152 (C4 (Web Text))

Max Activating Token Index: 656

Text #5

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.7507. Min Act: -0.1700

Data Index: 9173 (C4 (Web Text))

Max Activating Token Index: 1000

Text #6

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.7194. Min Act: -0.1700

Data Index: 757395 (C4 (Web Text))

Max Activating Token Index: 630

Text #7

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.6568. Min Act: -0.1700

Data Index: 1588826 (Python Code)

Max Activating Token Index: 31

Text #8

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.6568. Min Act: -0.1700

Data Index: 640313 (C4 (Web Text))

Max Activating Token Index: 788

Text #9

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.6568. Min Act: -0.1700

Data Index: 488061 (C4 (Web Text))

Max Activating Token Index: 268

Text #10

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.6568. Min Act: -0.1700

Data Index: 253280 (C4 (Web Text))

Max Activating Token Index: 767

Text #11

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.5942. Min Act: -0.1700

Data Index: 1574522 (Python Code)

Max Activating Token Index: 11

Text #12

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.6255. Min Act: -0.1699

Data Index: 1136146 (C4 (Web Text))

Max Activating Token Index: 922

Text #13

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.6255. Min Act: -0.1700

Data Index: 1169877 (C4 (Web Text))

Max Activating Token Index: 94

Text #14

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.5942. Min Act: -0.1700

Data Index: 1188859 (C4 (Web Text))

Max Activating Token Index: 566

Text #15

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.6255. Min Act: -0.1700

Data Index: 70696 (C4 (Web Text))

Max Activating Token Index: 711

Text #16

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.5942. Min Act: -0.1700

Data Index: 256620 (C4 (Web Text))

Max Activating Token Index: 706

Text #17

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.5942. Min Act: -0.1700

Data Index: 107232 (C4 (Web Text))

Max Activating Token Index: 634

Text #18

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.6568. Min Act: -0.1700

Data Index: 50659 (C4 (Web Text))

Max Activating Token Index: 777

Text #19

Max Range: 3.9071. Min Range: -3.9071

Max Act: 3.5942. Min Act: -0.1700

Data Index: 546740 (C4 (Web Text))

Max Activating Token Index: 636