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

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

Neuron 1765 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: 5.3944. Min Range: -5.3944

Max Act: 5.3944. Min Act: -0.1699

Data Index: 1630528 (Python Code)

Max Activating Token Index: 944

Text #1

Max Range: 5.3944. Min Range: -5.3944

Max Act: 5.3319. Min Act: -0.1700

Data Index: 1441543 (Python Code)

Max Activating Token Index: 792

Text #2

Max Range: 5.3944. Min Range: -5.3944

Max Act: 5.3007. Min Act: -0.1700

Data Index: 1631787 (Python Code)

Max Activating Token Index: 586

Text #3

Max Range: 5.3944. Min Range: -5.3944

Max Act: 5.1444. Min Act: -0.1700

Data Index: 1651243 (Python Code)

Max Activating Token Index: 644

Text #4

Max Range: 5.3944. Min Range: -5.3944

Max Act: 5.1757. Min Act: -0.1700

Data Index: 1631201 (Python Code)

Max Activating Token Index: 117

Text #5

Max Range: 5.3944. Min Range: -5.3944

Max Act: 5.0819. Min Act: -0.1700

Data Index: 1422587 (Python Code)

Max Activating Token Index: 1018

Text #6

Max Range: 5.3944. Min Range: -5.3944

Max Act: 5.0507. Min Act: -0.1700

Data Index: 1561041 (Python Code)

Max Activating Token Index: 556

Text #7

Max Range: 5.3944. Min Range: -5.3944

Max Act: 5.0507. Min Act: -0.1699

Data Index: 1549076 (Python Code)

Max Activating Token Index: 810

Text #8

Max Range: 5.3944. Min Range: -5.3944

Max Act: 5.0194. Min Act: -0.1700

Data Index: 1590251 (Python Code)

Max Activating Token Index: 714

Text #9

Max Range: 5.3944. Min Range: -5.3944

Max Act: 4.9569. Min Act: -0.1700

Data Index: 1379747 (Python Code)

Max Activating Token Index: 36

Text #10

Max Range: 5.3944. Min Range: -5.3944

Max Act: 4.9257. Min Act: -0.1700

Data Index: 1408987 (Python Code)

Max Activating Token Index: 93

Text #11

Max Range: 5.3944. Min Range: -5.3944

Max Act: 4.8944. Min Act: -0.1700

Data Index: 1197494 (C4 (Web Text))

Max Activating Token Index: 603

Text #12

Max Range: 5.3944. Min Range: -5.3944

Max Act: 4.8944. Min Act: -0.1700

Data Index: 1282904 (C4 (Web Text))

Max Activating Token Index: 665

Text #13

Max Range: 5.3944. Min Range: -5.3944

Max Act: 4.9257. Min Act: -0.1700

Data Index: 1448092 (Python Code)

Max Activating Token Index: 286

Text #14

Max Range: 5.3944. Min Range: -5.3944

Max Act: 4.9257. Min Act: -0.1700

Data Index: 1427658 (Python Code)

Max Activating Token Index: 342

Text #15

Max Range: 5.3944. Min Range: -5.3944

Max Act: 4.9257. Min Act: -0.1700

Data Index: 1405606 (Python Code)

Max Activating Token Index: 149

Text #16

Max Range: 5.3944. Min Range: -5.3944

Max Act: 4.8632. Min Act: -0.1700

Data Index: 1561390 (Python Code)

Max Activating Token Index: 880

Text #17

Max Range: 5.3944. Min Range: -5.3944

Max Act: 4.9257. Min Act: -0.1699

Data Index: 1619776 (Python Code)

Max Activating Token Index: 760

Text #18

Max Range: 5.3944. Min Range: -5.3944

Max Act: 4.8944. Min Act: -0.1700

Data Index: 1548728 (Python Code)

Max Activating Token Index: 188

Text #19

Max Range: 5.3944. Min Range: -5.3944

Max Act: 4.9257. Min Act: -0.1700

Data Index: 1450386 (Python Code)

Max Activating Token Index: 125