Model: GELU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 1260 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-3l')

Text #0

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.8331. Min Act: -0.1700

Data Index: 1132988 (C4 (Web Text))

Max Activating Token Index: 574

Text #1

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.7706. Min Act: -0.1700

Data Index: 178126 (C4 (Web Text))

Max Activating Token Index: 433

Text #2

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.7706. Min Act: -0.1700

Data Index: 1279389 (C4 (Web Text))

Max Activating Token Index: 719

Text #3

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.7080. Min Act: -0.1700

Data Index: 1415645 (Python Code)

Max Activating Token Index: 895

Text #4

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.7706. Min Act: -0.1700

Data Index: 1396610 (Python Code)

Max Activating Token Index: 768

Text #5

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.7393. Min Act: -0.1700

Data Index: 1359889 (Python Code)

Max Activating Token Index: 849

Text #6

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.7080. Min Act: -0.1700

Data Index: 1650100 (Python Code)

Max Activating Token Index: 757

Text #7

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.7080. Min Act: -0.1700

Data Index: 1591834 (Python Code)

Max Activating Token Index: 321

Text #8

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.7080. Min Act: -0.1700

Data Index: 1052935 (C4 (Web Text))

Max Activating Token Index: 1021

Text #9

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.7080. Min Act: -0.1700

Data Index: 456877 (C4 (Web Text))

Max Activating Token Index: 950

Text #10

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.6767. Min Act: -0.1700

Data Index: 920790 (C4 (Web Text))

Max Activating Token Index: 144

Text #11

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.6767. Min Act: -0.1700

Data Index: 1493613 (Python Code)

Max Activating Token Index: 940

Text #12

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.6767. Min Act: -0.1700

Data Index: 1539148 (Python Code)

Max Activating Token Index: 876

Text #13

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.6767. Min Act: -0.1700

Data Index: 1401014 (Python Code)

Max Activating Token Index: 694

Text #14

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.6767. Min Act: -0.1700

Data Index: 24257 (C4 (Web Text))

Max Activating Token Index: 712

Text #15

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.6767. Min Act: -0.1700

Data Index: 116944 (C4 (Web Text))

Max Activating Token Index: 532

Text #16

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.6141. Min Act: -0.1700

Data Index: 115588 (C4 (Web Text))

Max Activating Token Index: 183

Text #17

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.6141. Min Act: -0.1700

Data Index: 1609556 (Python Code)

Max Activating Token Index: 377

Text #18

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.6454. Min Act: -0.1700

Data Index: 1616560 (Python Code)

Max Activating Token Index: 963

Text #19

Max Range: 3.8331. Min Range: -3.8331

Max Act: 3.5828. Min Act: -0.1700

Data Index: 1471814 (Python Code)

Max Activating Token Index: 762