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

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

Neuron 390 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.6088. Min Act: -0.1700

Data Index: 832063 (C4 (Web Text))

Max Activating Token Index: 909

Text #1

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.5461. Min Act: -0.1700

Data Index: 1548656 (Python Code)

Max Activating Token Index: 560

Text #2

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.4835. Min Act: -0.1700

Data Index: 673175 (C4 (Web Text))

Max Activating Token Index: 358

Text #3

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.3580. Min Act: -0.1700

Data Index: 125576 (C4 (Web Text))

Max Activating Token Index: 333

Text #4

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.3580. Min Act: -0.1700

Data Index: 59425 (C4 (Web Text))

Max Activating Token Index: 609

Text #5

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.3266. Min Act: -0.1700

Data Index: 971488 (C4 (Web Text))

Max Activating Token Index: 389

Text #6

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.2638. Min Act: -0.1700

Data Index: 736134 (C4 (Web Text))

Max Activating Token Index: 571

Text #7

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.3266. Min Act: -0.1700

Data Index: 1277032 (C4 (Web Text))

Max Activating Token Index: 778

Text #8

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.1379. Min Act: -0.1700

Data Index: 336022 (C4 (Web Text))

Max Activating Token Index: 867

Text #9

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.2009. Min Act: -0.1700

Data Index: 156326 (C4 (Web Text))

Max Activating Token Index: 915

Text #10

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.2324. Min Act: -0.1699

Data Index: 1241624 (C4 (Web Text))

Max Activating Token Index: 620

Text #11

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.2009. Min Act: -0.1699

Data Index: 589988 (C4 (Web Text))

Max Activating Token Index: 156

Text #12

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.2009. Min Act: -0.1700

Data Index: 422216 (C4 (Web Text))

Max Activating Token Index: 298

Text #13

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.2009. Min Act: -0.1700

Data Index: 1074675 (C4 (Web Text))

Max Activating Token Index: 87

Text #14

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.2952. Min Act: -0.1700

Data Index: 1202724 (C4 (Web Text))

Max Activating Token Index: 213

Text #15

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.1694. Min Act: -0.1700

Data Index: 496356 (C4 (Web Text))

Max Activating Token Index: 434

Text #16

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.1694. Min Act: -0.1700

Data Index: 257189 (C4 (Web Text))

Max Activating Token Index: 919

Text #17

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.1379. Min Act: -0.1700

Data Index: 375828 (C4 (Web Text))

Max Activating Token Index: 912

Text #18

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.1379. Min Act: -0.1700

Data Index: 683518 (C4 (Web Text))

Max Activating Token Index: 398

Text #19

Max Range: 3.6088. Min Range: -3.6088

Max Act: 3.1379. Min Act: -0.1700

Data Index: 149833 (C4 (Web Text))

Max Activating Token Index: 569