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

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

Neuron 2025 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: 3.8762. Min Range: -3.8762

Max Act: 3.8762. Min Act: -0.1700

Data Index: 340803 (C4 (Web Text))

Max Activating Token Index: 676

Text #1

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.8450. Min Act: -0.1700

Data Index: 17664 (C4 (Web Text))

Max Activating Token Index: 648

Text #2

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.7667. Min Act: -0.1700

Data Index: 1306547 (C4 (Web Text))

Max Activating Token Index: 618

Text #3

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.7198. Min Act: -0.1700

Data Index: 413701 (C4 (Web Text))

Max Activating Token Index: 699

Text #4

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.7198. Min Act: -0.1700

Data Index: 603263 (C4 (Web Text))

Max Activating Token Index: 730

Text #5

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.7511. Min Act: -0.1700

Data Index: 424030 (C4 (Web Text))

Max Activating Token Index: 546

Text #6

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.6416. Min Act: -0.1700

Data Index: 1030885 (C4 (Web Text))

Max Activating Token Index: 841

Text #7

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.7355. Min Act: -0.1700

Data Index: 694815 (C4 (Web Text))

Max Activating Token Index: 449

Text #8

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.7042. Min Act: -0.1700

Data Index: 1197739 (C4 (Web Text))

Max Activating Token Index: 808

Text #9

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.7198. Min Act: -0.1700

Data Index: 228013 (C4 (Web Text))

Max Activating Token Index: 583

Text #10

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.6885. Min Act: -0.1700

Data Index: 1224972 (C4 (Web Text))

Max Activating Token Index: 575

Text #11

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.7042. Min Act: -0.1700

Data Index: 211029 (C4 (Web Text))

Max Activating Token Index: 882

Text #12

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.7042. Min Act: -0.1700

Data Index: 418280 (C4 (Web Text))

Max Activating Token Index: 435

Text #13

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.6885. Min Act: -0.1700

Data Index: 38537 (C4 (Web Text))

Max Activating Token Index: 181

Text #14

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.6103. Min Act: -0.1700

Data Index: 117348 (C4 (Web Text))

Max Activating Token Index: 442

Text #15

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.6416. Min Act: -0.1700

Data Index: 56206 (C4 (Web Text))

Max Activating Token Index: 668

Text #16

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.6416. Min Act: -0.1700

Data Index: 1156371 (C4 (Web Text))

Max Activating Token Index: 974

Text #17

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.5476. Min Act: -0.1700

Data Index: 233747 (C4 (Web Text))

Max Activating Token Index: 815

Text #18

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.5633. Min Act: -0.1700

Data Index: 1111747 (C4 (Web Text))

Max Activating Token Index: 1007

Text #19

Max Range: 3.8762. Min Range: -3.8762

Max Act: 3.5789. Min Act: -0.1700

Data Index: 585272 (C4 (Web Text))

Max Activating Token Index: 733