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

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

Neuron 1530 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.2762. Min Range: -5.2762

Max Act: 5.2762. Min Act: -0.1700

Data Index: 649745 (C4 (Web Text))

Max Activating Token Index: 760

Text #1

Max Range: 5.2762. Min Range: -5.2762

Max Act: 5.0887. Min Act: -0.1700

Data Index: 1004612 (C4 (Web Text))

Max Activating Token Index: 408

Text #2

Max Range: 5.2762. Min Range: -5.2762

Max Act: 5.0262. Min Act: -0.1700

Data Index: 1447621 (Python Code)

Max Activating Token Index: 543

Text #3

Max Range: 5.2762. Min Range: -5.2762

Max Act: 5.0262. Min Act: -0.1697

Data Index: 417110 (C4 (Web Text))

Max Activating Token Index: 168

Text #4

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.9949. Min Act: -0.1700

Data Index: 243783 (C4 (Web Text))

Max Activating Token Index: 537

Text #5

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.9637. Min Act: -0.1700

Data Index: 1634817 (Python Code)

Max Activating Token Index: 1005

Text #6

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.9949. Min Act: -0.1700

Data Index: 1446430 (Python Code)

Max Activating Token Index: 758

Text #7

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.9949. Min Act: -0.1700

Data Index: 949607 (C4 (Web Text))

Max Activating Token Index: 925

Text #8

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.9324. Min Act: -0.1700

Data Index: 967856 (C4 (Web Text))

Max Activating Token Index: 556

Text #9

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.9012. Min Act: -0.1700

Data Index: 930799 (C4 (Web Text))

Max Activating Token Index: 502

Text #10

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.8699. Min Act: -0.1700

Data Index: 1425816 (Python Code)

Max Activating Token Index: 722

Text #11

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.8699. Min Act: -0.1700

Data Index: 1088034 (C4 (Web Text))

Max Activating Token Index: 265

Text #12

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.8856. Min Act: -0.1700

Data Index: 1247080 (C4 (Web Text))

Max Activating Token Index: 511

Text #13

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.7918. Min Act: -0.1700

Data Index: 1544218 (Python Code)

Max Activating Token Index: 799

Text #14

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.8231. Min Act: -0.1700

Data Index: 1494603 (Python Code)

Max Activating Token Index: 520

Text #15

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.7918. Min Act: -0.1700

Data Index: 1371050 (Python Code)

Max Activating Token Index: 424

Text #16

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.8074. Min Act: -0.1700

Data Index: 822275 (C4 (Web Text))

Max Activating Token Index: 149

Text #17

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.8074. Min Act: -0.1700

Data Index: 33920 (C4 (Web Text))

Max Activating Token Index: 650

Text #18

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.7918. Min Act: -0.1700

Data Index: 743852 (C4 (Web Text))

Max Activating Token Index: 798

Text #19

Max Range: 5.2762. Min Range: -5.2762

Max Act: 4.7762. Min Act: -0.1700

Data Index: 1289169 (C4 (Web Text))

Max Activating Token Index: 71