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

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

Neuron 1060 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: 3.3216. Min Range: -3.3216

Max Act: 3.3216. Min Act: -0.1700

Data Index: 1615418 (Python Code)

Max Activating Token Index: 453

Text #1

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.2744. Min Act: -0.1700

Data Index: 1392061 (Python Code)

Max Activating Token Index: 467

Text #2

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.2587. Min Act: -0.1700

Data Index: 1439774 (Python Code)

Max Activating Token Index: 156

Text #3

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.2587. Min Act: -0.1700

Data Index: 1535638 (Python Code)

Max Activating Token Index: 306

Text #4

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.2273. Min Act: -0.1700

Data Index: 1619503 (Python Code)

Max Activating Token Index: 597

Text #5

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.2273. Min Act: -0.1700

Data Index: 1314245 (C4 (Web Text))

Max Activating Token Index: 649

Text #6

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.2273. Min Act: -0.1700

Data Index: 1510369 (Python Code)

Max Activating Token Index: 474

Text #7

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.2430. Min Act: -0.1700

Data Index: 1162547 (C4 (Web Text))

Max Activating Token Index: 749

Text #8

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.2116. Min Act: -0.1700

Data Index: 1498724 (Python Code)

Max Activating Token Index: 545

Text #9

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.2116. Min Act: -0.1700

Data Index: 972947 (C4 (Web Text))

Max Activating Token Index: 235

Text #10

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.2116. Min Act: -0.1700

Data Index: 541580 (C4 (Web Text))

Max Activating Token Index: 784

Text #11

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.2116. Min Act: -0.1700

Data Index: 756308 (C4 (Web Text))

Max Activating Token Index: 1011

Text #12

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.1801. Min Act: -0.1700

Data Index: 1268898 (C4 (Web Text))

Max Activating Token Index: 549

Text #13

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.2273. Min Act: -0.1700

Data Index: 1589549 (Python Code)

Max Activating Token Index: 454

Text #14

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.1958. Min Act: -0.1700

Data Index: 1502467 (Python Code)

Max Activating Token Index: 1023

Text #15

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.1958. Min Act: -0.1700

Data Index: 956546 (C4 (Web Text))

Max Activating Token Index: 390

Text #16

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.2116. Min Act: -0.1700

Data Index: 563760 (C4 (Web Text))

Max Activating Token Index: 924

Text #17

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.1801. Min Act: -0.1700

Data Index: 1025283 (C4 (Web Text))

Max Activating Token Index: 929

Text #18

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.1958. Min Act: -0.1700

Data Index: 1649254 (Python Code)

Max Activating Token Index: 279

Text #19

Max Range: 3.3216. Min Range: -3.3216

Max Act: 3.1644. Min Act: -0.1700

Data Index: 985971 (C4 (Web Text))

Max Activating Token Index: 581