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

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

Neuron 1265 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-3l')

Text #0

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.3323. Min Act: -0.1700

Data Index: 273780 (C4 (Web Text))

Max Activating Token Index: 38

Text #1

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.2698. Min Act: -0.1700

Data Index: 52532 (C4 (Web Text))

Max Activating Token Index: 575

Text #2

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.2698. Min Act: -0.1700

Data Index: 1357222 (C4 (Web Text))

Max Activating Token Index: 849

Text #3

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.2698. Min Act: -0.1700

Data Index: 1041771 (C4 (Web Text))

Max Activating Token Index: 587

Text #4

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.0823. Min Act: -0.1700

Data Index: 276549 (C4 (Web Text))

Max Activating Token Index: 786

Text #5

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.2698. Min Act: -0.1700

Data Index: 1250593 (C4 (Web Text))

Max Activating Token Index: 775

Text #6

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.2073. Min Act: -0.1700

Data Index: 227823 (C4 (Web Text))

Max Activating Token Index: 597

Text #7

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.1448. Min Act: -0.1700

Data Index: 550173 (C4 (Web Text))

Max Activating Token Index: 881

Text #8

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.1448. Min Act: -0.1700

Data Index: 776803 (C4 (Web Text))

Max Activating Token Index: 213

Text #9

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.1448. Min Act: -0.1700

Data Index: 869196 (C4 (Web Text))

Max Activating Token Index: 73

Text #10

Max Range: 8.3323. Min Range: -8.3323

Max Act: 7.8948. Min Act: -0.1699

Data Index: 605026 (C4 (Web Text))

Max Activating Token Index: 500

Text #11

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.0511. Min Act: -0.1700

Data Index: 1235435 (C4 (Web Text))

Max Activating Token Index: 427

Text #12

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.0823. Min Act: -0.1700

Data Index: 683655 (C4 (Web Text))

Max Activating Token Index: 500

Text #13

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.0823. Min Act: -0.1700

Data Index: 19001 (C4 (Web Text))

Max Activating Token Index: 308

Text #14

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.0823. Min Act: -0.1700

Data Index: 26392 (C4 (Web Text))

Max Activating Token Index: 465

Text #15

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.0511. Min Act: -0.1700

Data Index: 124872 (C4 (Web Text))

Max Activating Token Index: 463

Text #16

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.0511. Min Act: -0.1700

Data Index: 386738 (C4 (Web Text))

Max Activating Token Index: 235

Text #17

Max Range: 8.3323. Min Range: -8.3323

Max Act: 8.0198. Min Act: -0.1700

Data Index: 896500 (C4 (Web Text))

Max Activating Token Index: 615

Text #18

Max Range: 8.3323. Min Range: -8.3323

Max Act: 7.9573. Min Act: -0.1700

Data Index: 336103 (C4 (Web Text))

Max Activating Token Index: 283

Text #19

Max Range: 8.3323. Min Range: -8.3323

Max Act: 7.9886. Min Act: -0.1700

Data Index: 331155 (C4 (Web Text))

Max Activating Token Index: 228