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

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

Neuron 36 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.6972. Min Act: -0.1700

Data Index: 172401 (C4 (Web Text))

Max Activating Token Index: 585

Text #1

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.6346. Min Act: -0.1700

Data Index: 1243223 (C4 (Web Text))

Max Activating Token Index: 431

Text #2

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.6033. Min Act: -0.1700

Data Index: 368875 (C4 (Web Text))

Max Activating Token Index: 533

Text #3

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.6033. Min Act: -0.1700

Data Index: 1262586 (C4 (Web Text))

Max Activating Token Index: 311

Text #4

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.5720. Min Act: -0.1700

Data Index: 934989 (C4 (Web Text))

Max Activating Token Index: 765

Text #5

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.5094. Min Act: -0.1700

Data Index: 1168878 (C4 (Web Text))

Max Activating Token Index: 495

Text #6

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.5094. Min Act: -0.1700

Data Index: 586841 (C4 (Web Text))

Max Activating Token Index: 613

Text #7

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.4780. Min Act: -0.1700

Data Index: 286435 (C4 (Web Text))

Max Activating Token Index: 729

Text #8

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.5094. Min Act: -0.1700

Data Index: 425290 (C4 (Web Text))

Max Activating Token Index: 1000

Text #9

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.4467. Min Act: -0.1700

Data Index: 1251078 (C4 (Web Text))

Max Activating Token Index: 958

Text #10

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.4467. Min Act: -0.1700

Data Index: 1202516 (C4 (Web Text))

Max Activating Token Index: 690

Text #11

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.3526. Min Act: -0.1700

Data Index: 903513 (C4 (Web Text))

Max Activating Token Index: 320

Text #12

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.4153. Min Act: -0.1700

Data Index: 1033993 (C4 (Web Text))

Max Activating Token Index: 591

Text #13

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.4153. Min Act: -0.1700

Data Index: 1339969 (C4 (Web Text))

Max Activating Token Index: 239

Text #14

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.4467. Min Act: -0.1700

Data Index: 897874 (C4 (Web Text))

Max Activating Token Index: 380

Text #15

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.4153. Min Act: -0.1700

Data Index: 1296468 (C4 (Web Text))

Max Activating Token Index: 289

Text #16

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.4153. Min Act: -0.1700

Data Index: 825818 (C4 (Web Text))

Max Activating Token Index: 81

Text #17

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.4153. Min Act: -0.1700

Data Index: 735460 (C4 (Web Text))

Max Activating Token Index: 577

Text #18

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.3526. Min Act: -0.1700

Data Index: 328030 (C4 (Web Text))

Max Activating Token Index: 701

Text #19

Max Range: 3.6972. Min Range: -3.6972

Max Act: 3.3839. Min Act: -0.1700

Data Index: 988308 (C4 (Web Text))

Max Activating Token Index: 481