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

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

Neuron 1512 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: 3.2128. Min Range: -3.2128

Max Act: 3.1970. Min Act: -0.1700

Data Index: 503798 (C4 (Web Text))

Max Activating Token Index: 419

Text #1

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.2128. Min Act: -0.1700

Data Index: 406585 (C4 (Web Text))

Max Activating Token Index: 323

Text #2

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.1813. Min Act: -0.1700

Data Index: 520263 (C4 (Web Text))

Max Activating Token Index: 717

Text #3

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.1970. Min Act: -0.1700

Data Index: 443599 (C4 (Web Text))

Max Activating Token Index: 20

Text #4

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.1813. Min Act: -0.1700

Data Index: 937765 (C4 (Web Text))

Max Activating Token Index: 714

Text #5

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.1813. Min Act: -0.1700

Data Index: 1228297 (C4 (Web Text))

Max Activating Token Index: 422

Text #6

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.1341. Min Act: -0.1700

Data Index: 460775 (C4 (Web Text))

Max Activating Token Index: 233

Text #7

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.1341. Min Act: -0.1700

Data Index: 986673 (C4 (Web Text))

Max Activating Token Index: 216

Text #8

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.1341. Min Act: -0.1700

Data Index: 1191060 (C4 (Web Text))

Max Activating Token Index: 412

Text #9

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.1183. Min Act: -0.1700

Data Index: 1184362 (C4 (Web Text))

Max Activating Token Index: 133

Text #10

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.1183. Min Act: -0.1700

Data Index: 241473 (C4 (Web Text))

Max Activating Token Index: 290

Text #11

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.1341. Min Act: -0.1700

Data Index: 319124 (C4 (Web Text))

Max Activating Token Index: 642

Text #12

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.1025. Min Act: -0.1700

Data Index: 467391 (C4 (Web Text))

Max Activating Token Index: 268

Text #13

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.0868. Min Act: -0.1700

Data Index: 825776 (C4 (Web Text))

Max Activating Token Index: 609

Text #14

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.0868. Min Act: -0.1700

Data Index: 905055 (C4 (Web Text))

Max Activating Token Index: 564

Text #15

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.0868. Min Act: -0.1700

Data Index: 1424223 (Python Code)

Max Activating Token Index: 3

Text #16

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.0710. Min Act: -0.1700

Data Index: 778542 (C4 (Web Text))

Max Activating Token Index: 678

Text #17

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.1341. Min Act: -0.1700

Data Index: 490461 (C4 (Web Text))

Max Activating Token Index: 656

Text #18

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.0710. Min Act: -0.1700

Data Index: 1016924 (C4 (Web Text))

Max Activating Token Index: 90

Text #19

Max Range: 3.2128. Min Range: -3.2128

Max Act: 3.1025. Min Act: -0.1700

Data Index: 290274 (C4 (Web Text))

Max Activating Token Index: 116