Model: GELU Model: 1 Layers, 2048 Neurons per Layer

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

Neuron 1004 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

Hooked Transformer Loading: HookedTransformer.from_pretrained('gelu-1l')

Click the toggle to see the full text!

Text #0

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.9686. Min Act: -0.1700

Data Index: 1606784 (Python Code)

Max Activating Token Index: 10

Text #1

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.7469. Min Act: -0.1700

Data Index: 965426 (C4 (Web Text))

Max Activating Token Index: 619

Text #2

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.7695. Min Act: -0.1700

Data Index: 268400 (C4 (Web Text))

Max Activating Token Index: 31

Text #3

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.6725. Min Act: -0.1608

Data Index: 1357911 (C4 (Web Text))

Max Activating Token Index: 458

Text #4

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.6803. Min Act: -0.1337

Data Index: 1297199 (C4 (Web Text))

Max Activating Token Index: 572

Text #5

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.6425. Min Act: -0.0782

Data Index: 1493825 (Python Code)

Max Activating Token Index: 77

Text #6

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.5896. Min Act: -0.1545

Data Index: 1625076 (Python Code)

Max Activating Token Index: 56

Text #7

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.6346. Min Act: -0.1700

Data Index: 719888 (C4 (Web Text))

Max Activating Token Index: 874

Text #8

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.6134. Min Act: -0.1700

Data Index: 1093754 (C4 (Web Text))

Max Activating Token Index: 899

Text #9

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.6141. Min Act: -0.1700

Data Index: 527838 (C4 (Web Text))

Max Activating Token Index: 911

Text #10

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.6083. Min Act: -0.1700

Data Index: 1200550 (C4 (Web Text))

Max Activating Token Index: 779

Text #11

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.6318. Min Act: -0.1582

Data Index: 164466 (C4 (Web Text))

Max Activating Token Index: 910

Text #12

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.5829. Min Act: -0.1529

Data Index: 614602 (C4 (Web Text))

Max Activating Token Index: 520

Text #13

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.5560. Min Act: -0.1693

Data Index: 1423797 (Python Code)

Max Activating Token Index: 619

Text #14

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.5607. Min Act: -0.1694

Data Index: 1509414 (Python Code)

Max Activating Token Index: 738

Text #15

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.5420. Min Act: -0.1699

Data Index: 1606491 (Python Code)

Max Activating Token Index: 50

Text #16

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.5852. Min Act: -0.1699

Data Index: 1606393 (Python Code)

Max Activating Token Index: 106

Text #17

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.5421. Min Act: -0.0082

Data Index: 71643 (C4 (Web Text))

Max Activating Token Index: 35

Text #18

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.5931. Min Act: -0.1700

Data Index: 860679 (C4 (Web Text))

Max Activating Token Index: 597

Text #19

Max Range: 5.9686. Min Range: -5.9686

Max Act: 5.5449. Min Act: -0.1171

Data Index: 20217 (C4 (Web Text))

Max Activating Token Index: 1023