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

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

Neuron 876 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: 3.4177. Min Range: -3.4177

Max Act: 3.4177. Min Act: -0.1699

Data Index: 52005 (C4 (Web Text))

Max Activating Token Index: 964

Text #1

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.3236. Min Act: -0.1700

Data Index: 256047 (C4 (Web Text))

Max Activating Token Index: 1016

Text #2

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.2765. Min Act: -0.1700

Data Index: 853035 (C4 (Web Text))

Max Activating Token Index: 540

Text #3

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.3236. Min Act: -0.1700

Data Index: 698162 (C4 (Web Text))

Max Activating Token Index: 574

Text #4

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.2922. Min Act: -0.1700

Data Index: 1621287 (Python Code)

Max Activating Token Index: 955

Text #5

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.2922. Min Act: -0.1700

Data Index: 472942 (C4 (Web Text))

Max Activating Token Index: 857

Text #6

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.2293. Min Act: -0.1700

Data Index: 1003491 (C4 (Web Text))

Max Activating Token Index: 242

Text #7

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.2608. Min Act: -0.1700

Data Index: 1632017 (Python Code)

Max Activating Token Index: 789

Text #8

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.1821. Min Act: -0.1700

Data Index: 400405 (C4 (Web Text))

Max Activating Token Index: 522

Text #9

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.2136. Min Act: -0.1700

Data Index: 916629 (C4 (Web Text))

Max Activating Token Index: 675

Text #10

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.1821. Min Act: -0.1700

Data Index: 1430676 (Python Code)

Max Activating Token Index: 762

Text #11

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.1821. Min Act: -0.1700

Data Index: 403820 (C4 (Web Text))

Max Activating Token Index: 648

Text #12

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.1979. Min Act: -0.1700

Data Index: 11992 (C4 (Web Text))

Max Activating Token Index: 568

Text #13

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.1664. Min Act: -0.1700

Data Index: 875627 (C4 (Web Text))

Max Activating Token Index: 495

Text #14

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.1821. Min Act: -0.1700

Data Index: 1546861 (Python Code)

Max Activating Token Index: 515

Text #15

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.1664. Min Act: -0.1700

Data Index: 1034829 (C4 (Web Text))

Max Activating Token Index: 722

Text #16

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.1506. Min Act: -0.1700

Data Index: 1213532 (C4 (Web Text))

Max Activating Token Index: 576

Text #17

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.1506. Min Act: -0.1700

Data Index: 1284699 (C4 (Web Text))

Max Activating Token Index: 950

Text #18

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.1821. Min Act: -0.1700

Data Index: 909781 (C4 (Web Text))

Max Activating Token Index: 943

Text #19

Max Range: 3.4177. Min Range: -3.4177

Max Act: 3.1506. Min Act: -0.1700

Data Index: 1505621 (Python Code)

Max Activating Token Index: 999