Model: GELU Model: 2 Layers, 2048 Neurons per Layer

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

Neuron 1491 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-2l')

Text #0

Max Range: 4.1375. Min Range: -4.1375

Max Act: 4.1063. Min Act: -0.1700

Data Index: 512313 (C4 (Web Text))

Max Activating Token Index: 868

Text #1

Max Range: 4.1375. Min Range: -4.1375

Max Act: 4.1063. Min Act: -0.1700

Data Index: 181418 (C4 (Web Text))

Max Activating Token Index: 801

Text #2

Max Range: 4.1375. Min Range: -4.1375

Max Act: 4.1375. Min Act: -0.1700

Data Index: 529508 (C4 (Web Text))

Max Activating Token Index: 738

Text #3

Max Range: 4.1375. Min Range: -4.1375

Max Act: 4.1375. Min Act: -0.1700

Data Index: 203148 (C4 (Web Text))

Max Activating Token Index: 506

Text #4

Max Range: 4.1375. Min Range: -4.1375

Max Act: 4.0125. Min Act: -0.1700

Data Index: 1526907 (Python Code)

Max Activating Token Index: 266

Text #5

Max Range: 4.1375. Min Range: -4.1375

Max Act: 3.9812. Min Act: -0.1700

Data Index: 785389 (C4 (Web Text))

Max Activating Token Index: 101

Text #6

Max Range: 4.1375. Min Range: -4.1375

Max Act: 4.0125. Min Act: -0.1700

Data Index: 907396 (C4 (Web Text))

Max Activating Token Index: 291

Text #7

Max Range: 4.1375. Min Range: -4.1375

Max Act: 3.9499. Min Act: -0.1700

Data Index: 969991 (C4 (Web Text))

Max Activating Token Index: 609

Text #8

Max Range: 4.1375. Min Range: -4.1375

Max Act: 3.9187. Min Act: -0.1700

Data Index: 232139 (C4 (Web Text))

Max Activating Token Index: 210

Text #9

Max Range: 4.1375. Min Range: -4.1375

Max Act: 3.9499. Min Act: -0.1700

Data Index: 1531411 (Python Code)

Max Activating Token Index: 578

Text #10

Max Range: 4.1375. Min Range: -4.1375

Max Act: 3.9499. Min Act: -0.1700

Data Index: 367435 (C4 (Web Text))

Max Activating Token Index: 742

Text #11

Max Range: 4.1375. Min Range: -4.1375

Max Act: 3.8874. Min Act: -0.1700

Data Index: 182148 (C4 (Web Text))

Max Activating Token Index: 931

Text #12

Max Range: 4.1375. Min Range: -4.1375

Max Act: 3.8874. Min Act: -0.1700

Data Index: 1527185 (Python Code)

Max Activating Token Index: 420

Text #13

Max Range: 4.1375. Min Range: -4.1375

Max Act: 3.8561. Min Act: -0.1700

Data Index: 734592 (C4 (Web Text))

Max Activating Token Index: 365

Text #14

Max Range: 4.1375. Min Range: -4.1375

Max Act: 3.8561. Min Act: -0.1700

Data Index: 810665 (C4 (Web Text))

Max Activating Token Index: 837

Text #15

Max Range: 4.1375. Min Range: -4.1375

Max Act: 3.7936. Min Act: -0.1700

Data Index: 1152772 (C4 (Web Text))

Max Activating Token Index: 917

Text #16

Max Range: 4.1375. Min Range: -4.1375

Max Act: 3.8248. Min Act: -0.1700

Data Index: 1477077 (Python Code)

Max Activating Token Index: 722

Text #17

Max Range: 4.1375. Min Range: -4.1375

Max Act: 3.8561. Min Act: -0.1700

Data Index: 1206325 (C4 (Web Text))

Max Activating Token Index: 118

Text #18

Max Range: 4.1375. Min Range: -4.1375

Max Act: 3.7936. Min Act: -0.1700

Data Index: 434548 (C4 (Web Text))

Max Activating Token Index: 24

Text #19

Max Range: 4.1375. Min Range: -4.1375

Max Act: 3.7936. Min Act: -0.1700

Data Index: 657322 (C4 (Web Text))

Max Activating Token Index: 837