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

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

Neuron 1021 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: 7.1503. Min Range: -7.1503

Max Act: 7.1503. Min Act: -0.1699

Data Index: 1430568 (Python Code)

Max Activating Token Index: 1005

Text #1

Max Range: 7.1503. Min Range: -7.1503

Max Act: 7.1503. Min Act: -0.1700

Data Index: 1450589 (Python Code)

Max Activating Token Index: 919

Text #2

Max Range: 7.1503. Min Range: -7.1503

Max Act: 7.1191. Min Act: -0.1700

Data Index: 1568369 (Python Code)

Max Activating Token Index: 809

Text #3

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.9628. Min Act: -0.1700

Data Index: 1572071 (Python Code)

Max Activating Token Index: 795

Text #4

Max Range: 7.1503. Min Range: -7.1503

Max Act: 7.0566. Min Act: -0.1700

Data Index: 1512546 (Python Code)

Max Activating Token Index: 1023

Text #5

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.9941. Min Act: -0.1700

Data Index: 1463564 (Python Code)

Max Activating Token Index: 917

Text #6

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.9941. Min Act: -0.1700

Data Index: 1583792 (Python Code)

Max Activating Token Index: 407

Text #7

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.8378. Min Act: -0.1700

Data Index: 1588818 (Python Code)

Max Activating Token Index: 818

Text #8

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.8378. Min Act: -0.1700

Data Index: 1547132 (Python Code)

Max Activating Token Index: 241

Text #9

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.8066. Min Act: -0.1700

Data Index: 1617647 (Python Code)

Max Activating Token Index: 803

Text #10

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.8691. Min Act: -0.1700

Data Index: 1528990 (Python Code)

Max Activating Token Index: 995

Text #11

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.7753. Min Act: -0.1700

Data Index: 1566439 (Python Code)

Max Activating Token Index: 1023

Text #12

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.7128. Min Act: -0.1698

Data Index: 1548315 (Python Code)

Max Activating Token Index: 1023

Text #13

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.8066. Min Act: -0.1700

Data Index: 1583825 (Python Code)

Max Activating Token Index: 738

Text #14

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.8378. Min Act: -0.1700

Data Index: 1574451 (Python Code)

Max Activating Token Index: 402

Text #15

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.7441. Min Act: -0.1700

Data Index: 1494198 (Python Code)

Max Activating Token Index: 750

Text #16

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.6816. Min Act: -0.1700

Data Index: 1528539 (Python Code)

Max Activating Token Index: 344

Text #17

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.7128. Min Act: -0.1700

Data Index: 1381725 (Python Code)

Max Activating Token Index: 857

Text #18

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.7441. Min Act: -0.1700

Data Index: 1565505 (Python Code)

Max Activating Token Index: 765

Text #19

Max Range: 7.1503. Min Range: -7.1503

Max Act: 6.7441. Min Act: -0.1700

Data Index: 1561925 (Python Code)

Max Activating Token Index: 853