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

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

Neuron 508 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.6392. Min Act: -0.1700

Data Index: 1614086 (Python Code)

Max Activating Token Index: 428

Text #1

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.6080. Min Act: -0.1700

Data Index: 1653059 (Python Code)

Max Activating Token Index: 11

Text #2

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.6080. Min Act: -0.1700

Data Index: 1640475 (Python Code)

Max Activating Token Index: 885

Text #3

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.4517. Min Act: -0.1700

Data Index: 1385210 (Python Code)

Max Activating Token Index: 846

Text #4

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.4517. Min Act: -0.1700

Data Index: 1575721 (Python Code)

Max Activating Token Index: 1000

Text #5

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.4517. Min Act: -0.1700

Data Index: 1418760 (Python Code)

Max Activating Token Index: 264

Text #6

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.4205. Min Act: -0.1700

Data Index: 1436073 (Python Code)

Max Activating Token Index: 403

Text #7

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.3892. Min Act: -0.1700

Data Index: 1582042 (Python Code)

Max Activating Token Index: 177

Text #8

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.2642. Min Act: -0.1700

Data Index: 1394118 (Python Code)

Max Activating Token Index: 297

Text #9

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.3580. Min Act: -0.1700

Data Index: 1577784 (Python Code)

Max Activating Token Index: 249

Text #10

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.3267. Min Act: -0.1700

Data Index: 1571864 (Python Code)

Max Activating Token Index: 1015

Text #11

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.3267. Min Act: -0.1700

Data Index: 1388419 (Python Code)

Max Activating Token Index: 343

Text #12

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.2642. Min Act: -0.1700

Data Index: 1585336 (Python Code)

Max Activating Token Index: 620

Text #13

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.2955. Min Act: -0.1700

Data Index: 1533288 (Python Code)

Max Activating Token Index: 476

Text #14

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.3267. Min Act: -0.1700

Data Index: 1552321 (Python Code)

Max Activating Token Index: 455

Text #15

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.2642. Min Act: -0.1700

Data Index: 1640673 (Python Code)

Max Activating Token Index: 905

Text #16

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.2330. Min Act: -0.1700

Data Index: 1465970 (Python Code)

Max Activating Token Index: 500

Text #17

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.2642. Min Act: -0.1700

Data Index: 1614729 (Python Code)

Max Activating Token Index: 340

Text #18

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.2330. Min Act: -0.1700

Data Index: 1363225 (Python Code)

Max Activating Token Index: 292

Text #19

Max Range: 5.6392. Min Range: -5.6392

Max Act: 5.2330. Min Act: -0.1700

Data Index: 1483505 (Python Code)

Max Activating Token Index: 175