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

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

Neuron 650 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: 3.6342. Min Range: -3.6342

Max Act: 3.6342. Min Act: -0.1700

Data Index: 208640 (C4 (Web Text))

Max Activating Token Index: 271

Text #1

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.5716. Min Act: -0.1700

Data Index: 607446 (C4 (Web Text))

Max Activating Token Index: 995

Text #2

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.5559. Min Act: -0.1700

Data Index: 593592 (C4 (Web Text))

Max Activating Token Index: 946

Text #3

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.5089. Min Act: -0.1700

Data Index: 241188 (C4 (Web Text))

Max Activating Token Index: 633

Text #4

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.5403. Min Act: -0.1700

Data Index: 907 (C4 (Web Text))

Max Activating Token Index: 367

Text #5

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.5089. Min Act: -0.1700

Data Index: 168818 (C4 (Web Text))

Max Activating Token Index: 911

Text #6

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.4776. Min Act: -0.1700

Data Index: 1221666 (C4 (Web Text))

Max Activating Token Index: 513

Text #7

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.4776. Min Act: -0.1700

Data Index: 913930 (C4 (Web Text))

Max Activating Token Index: 559

Text #8

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.4619. Min Act: -0.1700

Data Index: 217084 (C4 (Web Text))

Max Activating Token Index: 315

Text #9

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.4462. Min Act: -0.1700

Data Index: 855653 (C4 (Web Text))

Max Activating Token Index: 192

Text #10

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.4619. Min Act: -0.1700

Data Index: 376028 (C4 (Web Text))

Max Activating Token Index: 161

Text #11

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.4933. Min Act: -0.1700

Data Index: 560200 (C4 (Web Text))

Max Activating Token Index: 268

Text #12

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.4619. Min Act: -0.1700

Data Index: 405013 (C4 (Web Text))

Max Activating Token Index: 887

Text #13

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.4776. Min Act: -0.1700

Data Index: 157504 (C4 (Web Text))

Max Activating Token Index: 184

Text #14

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.4619. Min Act: -0.1700

Data Index: 237 (C4 (Web Text))

Max Activating Token Index: 609

Text #15

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.4306. Min Act: -0.1700

Data Index: 742323 (C4 (Web Text))

Max Activating Token Index: 932

Text #16

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.4149. Min Act: -0.1700

Data Index: 197229 (C4 (Web Text))

Max Activating Token Index: 387

Text #17

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.4462. Min Act: -0.1700

Data Index: 777373 (C4 (Web Text))

Max Activating Token Index: 49

Text #18

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.4776. Min Act: -0.1700

Data Index: 193967 (C4 (Web Text))

Max Activating Token Index: 211

Text #19

Max Range: 3.6342. Min Range: -3.6342

Max Act: 3.4306. Min Act: -0.1700

Data Index: 1026755 (C4 (Web Text))

Max Activating Token Index: 76