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

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

Neuron 86 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.2180. Min Act: -0.1700

Data Index: 21263 (C4 (Web Text))

Max Activating Token Index: 216

Text #1

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.0920. Min Act: -0.1700

Data Index: 1384535 (Python Code)

Max Activating Token Index: 311

Text #2

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.0920. Min Act: -0.1700

Data Index: 1062818 (C4 (Web Text))

Max Activating Token Index: 699

Text #3

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.0920. Min Act: -0.1700

Data Index: 793776 (C4 (Web Text))

Max Activating Token Index: 268

Text #4

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.0920. Min Act: -0.1700

Data Index: 1335859 (C4 (Web Text))

Max Activating Token Index: 277

Text #5

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.1550. Min Act: -0.1700

Data Index: 319407 (C4 (Web Text))

Max Activating Token Index: 447

Text #6

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.1078. Min Act: -0.1700

Data Index: 1159258 (C4 (Web Text))

Max Activating Token Index: 432

Text #7

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.0604. Min Act: -0.1700

Data Index: 1113779 (C4 (Web Text))

Max Activating Token Index: 543

Text #8

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.0604. Min Act: -0.1700

Data Index: 63570 (C4 (Web Text))

Max Activating Token Index: 549

Text #9

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.0289. Min Act: -0.1700

Data Index: 1058112 (C4 (Web Text))

Max Activating Token Index: 438

Text #10

Max Range: 3.2180. Min Range: -3.2180

Max Act: 2.9973. Min Act: -0.1700

Data Index: 1564153 (Python Code)

Max Activating Token Index: 217

Text #11

Max Range: 3.2180. Min Range: -3.2180

Max Act: 2.9498. Min Act: -0.1700

Data Index: 1521048 (Python Code)

Max Activating Token Index: 412

Text #12

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.0289. Min Act: -0.1700

Data Index: 392556 (C4 (Web Text))

Max Activating Token Index: 212

Text #13

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.0131. Min Act: -0.1700

Data Index: 37728 (C4 (Web Text))

Max Activating Token Index: 825

Text #14

Max Range: 3.2180. Min Range: -3.2180

Max Act: 2.9814. Min Act: -0.1700

Data Index: 1268713 (C4 (Web Text))

Max Activating Token Index: 708

Text #15

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.0131. Min Act: -0.1700

Data Index: 1534762 (Python Code)

Max Activating Token Index: 227

Text #16

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.0447. Min Act: -0.1700

Data Index: 276279 (C4 (Web Text))

Max Activating Token Index: 133

Text #17

Max Range: 3.2180. Min Range: -3.2180

Max Act: 2.9498. Min Act: -0.1700

Data Index: 1157514 (C4 (Web Text))

Max Activating Token Index: 576

Text #18

Max Range: 3.2180. Min Range: -3.2180

Max Act: 2.9973. Min Act: -0.1700

Data Index: 317156 (C4 (Web Text))

Max Activating Token Index: 175

Text #19

Max Range: 3.2180. Min Range: -3.2180

Max Act: 3.0131. Min Act: -0.1700

Data Index: 297456 (C4 (Web Text))

Max Activating Token Index: 930