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

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

Neuron 882 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: 8.6009. Min Range: -8.6009

Max Act: 8.6009. Min Act: -0.1700

Data Index: 1482861 (Python Code)

Max Activating Token Index: 219

Text #1

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.5384. Min Act: -0.1700

Data Index: 430106 (C4 (Web Text))

Max Activating Token Index: 745

Text #2

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.4759. Min Act: -0.1700

Data Index: 1403873 (Python Code)

Max Activating Token Index: 522

Text #3

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.5384. Min Act: -0.1700

Data Index: 1414310 (Python Code)

Max Activating Token Index: 827

Text #4

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.2884. Min Act: -0.1700

Data Index: 1523120 (Python Code)

Max Activating Token Index: 227

Text #5

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.3197. Min Act: -0.1700

Data Index: 1379863 (Python Code)

Max Activating Token Index: 853

Text #6

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.3197. Min Act: -0.1700

Data Index: 1655784 (Python Code)

Max Activating Token Index: 295

Text #7

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.3509. Min Act: -0.1700

Data Index: 1651763 (Python Code)

Max Activating Token Index: 487

Text #8

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.2884. Min Act: -0.1700

Data Index: 1538054 (Python Code)

Max Activating Token Index: 578

Text #9

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.2884. Min Act: -0.1700

Data Index: 1459148 (Python Code)

Max Activating Token Index: 139

Text #10

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.2884. Min Act: -0.1700

Data Index: 1632561 (Python Code)

Max Activating Token Index: 578

Text #11

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.2259. Min Act: -0.1700

Data Index: 1582600 (Python Code)

Max Activating Token Index: 218

Text #12

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.1947. Min Act: -0.1700

Data Index: 1529537 (Python Code)

Max Activating Token Index: 293

Text #13

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.3822. Min Act: -0.1700

Data Index: 1651611 (Python Code)

Max Activating Token Index: 393

Text #14

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.1947. Min Act: -0.1700

Data Index: 1385094 (Python Code)

Max Activating Token Index: 916

Text #15

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.1634. Min Act: -0.1700

Data Index: 1395472 (Python Code)

Max Activating Token Index: 554

Text #16

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.1009. Min Act: -0.1700

Data Index: 1633009 (Python Code)

Max Activating Token Index: 1001

Text #17

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.1009. Min Act: -0.1700

Data Index: 222315 (C4 (Web Text))

Max Activating Token Index: 716

Text #18

Max Range: 8.6009. Min Range: -8.6009

Max Act: 7.9759. Min Act: -0.1700

Data Index: 1512520 (Python Code)

Max Activating Token Index: 261

Text #19

Max Range: 8.6009. Min Range: -8.6009

Max Act: 8.0072. Min Act: -0.1700

Data Index: 1439278 (Python Code)

Max Activating Token Index: 760