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

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

Neuron 1065 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: 3.9935. Min Range: -3.9935

Max Act: 3.9935. Min Act: -0.1700

Data Index: 344777 (C4 (Web Text))

Max Activating Token Index: 927

Text #1

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.9623. Min Act: -0.1700

Data Index: 563836 (C4 (Web Text))

Max Activating Token Index: 1019

Text #2

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.9623. Min Act: -0.1700

Data Index: 1374674 (Python Code)

Max Activating Token Index: 929

Text #3

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.8997. Min Act: -0.1700

Data Index: 841677 (C4 (Web Text))

Max Activating Token Index: 549

Text #4

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.9310. Min Act: -0.1700

Data Index: 48762 (C4 (Web Text))

Max Activating Token Index: 860

Text #5

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.9310. Min Act: -0.1700

Data Index: 725660 (C4 (Web Text))

Max Activating Token Index: 377

Text #6

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.8997. Min Act: -0.1700

Data Index: 711790 (C4 (Web Text))

Max Activating Token Index: 155

Text #7

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.8372. Min Act: -0.1700

Data Index: 1475539 (Python Code)

Max Activating Token Index: 33

Text #8

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.8372. Min Act: -0.1700

Data Index: 334878 (C4 (Web Text))

Max Activating Token Index: 514

Text #9

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.8372. Min Act: -0.1700

Data Index: 684541 (C4 (Web Text))

Max Activating Token Index: 902

Text #10

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.8059. Min Act: -0.1700

Data Index: 539388 (C4 (Web Text))

Max Activating Token Index: 833

Text #11

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.8059. Min Act: -0.1700

Data Index: 51427 (C4 (Web Text))

Max Activating Token Index: 473

Text #12

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.8059. Min Act: -0.1700

Data Index: 1306779 (C4 (Web Text))

Max Activating Token Index: 164

Text #13

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.8372. Min Act: -0.1700

Data Index: 443754 (C4 (Web Text))

Max Activating Token Index: 472

Text #14

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.8372. Min Act: -0.1700

Data Index: 861866 (C4 (Web Text))

Max Activating Token Index: 342

Text #15

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.8059. Min Act: -0.1700

Data Index: 469822 (C4 (Web Text))

Max Activating Token Index: 611

Text #16

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.8059. Min Act: -0.1700

Data Index: 1171814 (C4 (Web Text))

Max Activating Token Index: 937

Text #17

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.8059. Min Act: -0.1700

Data Index: 786716 (C4 (Web Text))

Max Activating Token Index: 516

Text #18

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.8059. Min Act: -0.1700

Data Index: 1606901 (Python Code)

Max Activating Token Index: 781

Text #19

Max Range: 3.9935. Min Range: -3.9935

Max Act: 3.7746. Min Act: -0.1700

Data Index: 1615936 (Python Code)

Max Activating Token Index: 20