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

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

Neuron 605 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.8074. Min Act: -0.1700

Data Index: 204541 (C4 (Web Text))

Max Activating Token Index: 358

Text #1

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.6824. Min Act: -0.1700

Data Index: 232391 (C4 (Web Text))

Max Activating Token Index: 450

Text #2

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.4949. Min Act: -0.1700

Data Index: 793415 (C4 (Web Text))

Max Activating Token Index: 731

Text #3

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.4949. Min Act: -0.1700

Data Index: 32305 (C4 (Web Text))

Max Activating Token Index: 313

Text #4

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.5574. Min Act: -0.1700

Data Index: 614469 (C4 (Web Text))

Max Activating Token Index: 897

Text #5

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.5574. Min Act: -0.1700

Data Index: 595601 (C4 (Web Text))

Max Activating Token Index: 911

Text #6

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.3699. Min Act: -0.1700

Data Index: 1189198 (C4 (Web Text))

Max Activating Token Index: 617

Text #7

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.3386. Min Act: -0.1700

Data Index: 862871 (C4 (Web Text))

Max Activating Token Index: 897

Text #8

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.4011. Min Act: -0.1700

Data Index: 1585878 (Python Code)

Max Activating Token Index: 829

Text #9

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.3699. Min Act: -0.1700

Data Index: 1251170 (C4 (Web Text))

Max Activating Token Index: 643

Text #10

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.3386. Min Act: -0.1700

Data Index: 834339 (C4 (Web Text))

Max Activating Token Index: 666

Text #11

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.3386. Min Act: -0.1700

Data Index: 349409 (C4 (Web Text))

Max Activating Token Index: 695

Text #12

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.3699. Min Act: -0.1700

Data Index: 1074982 (C4 (Web Text))

Max Activating Token Index: 918

Text #13

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.3074. Min Act: -0.1700

Data Index: 1484562 (Python Code)

Max Activating Token Index: 372

Text #14

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.2449. Min Act: -0.1700

Data Index: 212051 (C4 (Web Text))

Max Activating Token Index: 798

Text #15

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.3074. Min Act: -0.1700

Data Index: 244037 (C4 (Web Text))

Max Activating Token Index: 875

Text #16

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.2449. Min Act: -0.1700

Data Index: 584027 (C4 (Web Text))

Max Activating Token Index: 993

Text #17

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.1824. Min Act: -0.1700

Data Index: 1198068 (C4 (Web Text))

Max Activating Token Index: 834

Text #18

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.2761. Min Act: -0.1700

Data Index: 646487 (C4 (Web Text))

Max Activating Token Index: 999

Text #19

Max Range: 5.8074. Min Range: -5.8074

Max Act: 5.2136. Min Act: -0.1700

Data Index: 776469 (C4 (Web Text))

Max Activating Token Index: 693