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

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

Neuron 1415 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-4l')

Text #0

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.5140. Min Act: -0.1700

Data Index: 703421 (C4 (Web Text))

Max Activating Token Index: 823

Text #1

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.4043. Min Act: -0.1700

Data Index: 1275651 (C4 (Web Text))

Max Activating Token Index: 216

Text #2

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3886. Min Act: -0.1700

Data Index: 1174942 (C4 (Web Text))

Max Activating Token Index: 767

Text #3

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3729. Min Act: -0.1700

Data Index: 210883 (C4 (Web Text))

Max Activating Token Index: 410

Text #4

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3886. Min Act: -0.1700

Data Index: 1263402 (C4 (Web Text))

Max Activating Token Index: 863

Text #5

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3729. Min Act: -0.1700

Data Index: 286395 (C4 (Web Text))

Max Activating Token Index: 57

Text #6

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3729. Min Act: -0.1700

Data Index: 9593 (C4 (Web Text))

Max Activating Token Index: 600

Text #7

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3572. Min Act: -0.1700

Data Index: 964860 (C4 (Web Text))

Max Activating Token Index: 910

Text #8

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3258. Min Act: -0.1700

Data Index: 1557959 (Python Code)

Max Activating Token Index: 238

Text #9

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3415. Min Act: -0.1700

Data Index: 962712 (C4 (Web Text))

Max Activating Token Index: 448

Text #10

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3415. Min Act: -0.1700

Data Index: 346163 (C4 (Web Text))

Max Activating Token Index: 843

Text #11

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3415. Min Act: -0.1700

Data Index: 1550713 (Python Code)

Max Activating Token Index: 562

Text #12

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3258. Min Act: -0.1700

Data Index: 1612599 (Python Code)

Max Activating Token Index: 374

Text #13

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3101. Min Act: -0.1700

Data Index: 1076734 (C4 (Web Text))

Max Activating Token Index: 321

Text #14

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3101. Min Act: -0.1700

Data Index: 1526467 (Python Code)

Max Activating Token Index: 737

Text #15

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.2944. Min Act: -0.1700

Data Index: 1536308 (Python Code)

Max Activating Token Index: 257

Text #16

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3101. Min Act: -0.1700

Data Index: 585601 (C4 (Web Text))

Max Activating Token Index: 598

Text #17

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.2944. Min Act: -0.1700

Data Index: 1001710 (C4 (Web Text))

Max Activating Token Index: 739

Text #18

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.3101. Min Act: -0.1700

Data Index: 151230 (C4 (Web Text))

Max Activating Token Index: 342

Text #19

Max Range: 3.5140. Min Range: -3.5140

Max Act: 3.2787. Min Act: -0.1700

Data Index: 332958 (C4 (Web Text))

Max Activating Token Index: 810