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

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

Neuron 2034 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: 7.1755. Min Range: -7.1755

Max Act: 7.1755. Min Act: -0.1700

Data Index: 1562960 (Python Code)

Max Activating Token Index: 604

Text #1

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.9255. Min Act: -0.1700

Data Index: 1394011 (Python Code)

Max Activating Token Index: 399

Text #2

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.7692. Min Act: -0.1699

Data Index: 1611684 (Python Code)

Max Activating Token Index: 399

Text #3

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.8942. Min Act: -0.1700

Data Index: 1570236 (Python Code)

Max Activating Token Index: 252

Text #4

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.6442. Min Act: -0.1700

Data Index: 1648210 (Python Code)

Max Activating Token Index: 716

Text #5

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.7380. Min Act: -0.1700

Data Index: 1427345 (Python Code)

Max Activating Token Index: 946

Text #6

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.7067. Min Act: -0.1700

Data Index: 1579812 (Python Code)

Max Activating Token Index: 207

Text #7

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.5817. Min Act: -0.1700

Data Index: 1519944 (Python Code)

Max Activating Token Index: 165

Text #8

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.5505. Min Act: -0.1700

Data Index: 1433208 (Python Code)

Max Activating Token Index: 702

Text #9

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.4880. Min Act: -0.1700

Data Index: 1411212 (Python Code)

Max Activating Token Index: 172

Text #10

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.4880. Min Act: -0.1700

Data Index: 1402905 (Python Code)

Max Activating Token Index: 394

Text #11

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.5192. Min Act: -0.1699

Data Index: 1487595 (Python Code)

Max Activating Token Index: 353

Text #12

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.5505. Min Act: -0.1700

Data Index: 1420804 (Python Code)

Max Activating Token Index: 558

Text #13

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.5817. Min Act: -0.1699

Data Index: 1564608 (Python Code)

Max Activating Token Index: 825

Text #14

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.5505. Min Act: -0.1700

Data Index: 1422708 (Python Code)

Max Activating Token Index: 988

Text #15

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.3317. Min Act: -0.1700

Data Index: 1635938 (Python Code)

Max Activating Token Index: 653

Text #16

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.4567. Min Act: -0.1699

Data Index: 1519424 (Python Code)

Max Activating Token Index: 908

Text #17

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.4255. Min Act: -0.1700

Data Index: 1564166 (Python Code)

Max Activating Token Index: 933

Text #18

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.3630. Min Act: -0.1700

Data Index: 1403946 (Python Code)

Max Activating Token Index: 744

Text #19

Max Range: 7.1755. Min Range: -7.1755

Max Act: 6.3942. Min Act: -0.1700

Data Index: 1441290 (Python Code)

Max Activating Token Index: 997