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

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

Neuron 1655 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: 3.9824. Min Range: -3.9824

Max Act: 3.9824. Min Act: -0.1700

Data Index: 314352 (C4 (Web Text))

Max Activating Token Index: 575

Text #1

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.8261. Min Act: -0.1700

Data Index: 647178 (C4 (Web Text))

Max Activating Token Index: 638

Text #2

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.7948. Min Act: -0.1700

Data Index: 183002 (C4 (Web Text))

Max Activating Token Index: 576

Text #3

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.7009. Min Act: -0.1700

Data Index: 110685 (C4 (Web Text))

Max Activating Token Index: 1020

Text #4

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.7009. Min Act: -0.1700

Data Index: 199183 (C4 (Web Text))

Max Activating Token Index: 640

Text #5

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.6383. Min Act: -0.1700

Data Index: 133182 (C4 (Web Text))

Max Activating Token Index: 583

Text #6

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.6383. Min Act: -0.1700

Data Index: 65236 (C4 (Web Text))

Max Activating Token Index: 817

Text #7

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.5757. Min Act: -0.1700

Data Index: 159048 (C4 (Web Text))

Max Activating Token Index: 477

Text #8

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.5600. Min Act: -0.1700

Data Index: 433759 (C4 (Web Text))

Max Activating Token Index: 278

Text #9

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.5444. Min Act: -0.1700

Data Index: 607889 (C4 (Web Text))

Max Activating Token Index: 283

Text #10

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.5287. Min Act: -0.1700

Data Index: 1433542 (Python Code)

Max Activating Token Index: 335

Text #11

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.5130. Min Act: -0.1700

Data Index: 1464471 (Python Code)

Max Activating Token Index: 620

Text #12

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.4974. Min Act: -0.1700

Data Index: 375546 (C4 (Web Text))

Max Activating Token Index: 475

Text #13

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.4974. Min Act: -0.1700

Data Index: 989743 (C4 (Web Text))

Max Activating Token Index: 958

Text #14

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.4817. Min Act: -0.1700

Data Index: 371244 (C4 (Web Text))

Max Activating Token Index: 735

Text #15

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.4974. Min Act: -0.1700

Data Index: 1523798 (Python Code)

Max Activating Token Index: 598

Text #16

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.4817. Min Act: -0.1700

Data Index: 404661 (C4 (Web Text))

Max Activating Token Index: 287

Text #17

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.4817. Min Act: -0.1700

Data Index: 1174138 (C4 (Web Text))

Max Activating Token Index: 844

Text #18

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.4503. Min Act: -0.1700

Data Index: 1518407 (Python Code)

Max Activating Token Index: 494

Text #19

Max Range: 3.9824. Min Range: -3.9824

Max Act: 3.4817. Min Act: -0.1700

Data Index: 1412636 (Python Code)

Max Activating Token Index: 702