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

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

Neuron 622 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.7900. Min Act: -0.1700

Data Index: 872189 (C4 (Web Text))

Max Activating Token Index: 1015

Text #1

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.8212. Min Act: -0.1700

Data Index: 6675 (C4 (Web Text))

Max Activating Token Index: 83

Text #2

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.7587. Min Act: -0.1700

Data Index: 85290 (C4 (Web Text))

Max Activating Token Index: 878

Text #3

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.6025. Min Act: -0.1700

Data Index: 198325 (C4 (Web Text))

Max Activating Token Index: 542

Text #4

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.5400. Min Act: -0.1700

Data Index: 539447 (C4 (Web Text))

Max Activating Token Index: 967

Text #5

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.4775. Min Act: -0.1697

Data Index: 168108 (C4 (Web Text))

Max Activating Token Index: 484

Text #6

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.5087. Min Act: -0.1700

Data Index: 522757 (C4 (Web Text))

Max Activating Token Index: 441

Text #7

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.5400. Min Act: -0.1700

Data Index: 1215948 (C4 (Web Text))

Max Activating Token Index: 736

Text #8

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.4775. Min Act: -0.1700

Data Index: 908203 (C4 (Web Text))

Max Activating Token Index: 1000

Text #9

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.4775. Min Act: -0.1699

Data Index: 344098 (C4 (Web Text))

Max Activating Token Index: 677

Text #10

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.4775. Min Act: -0.1700

Data Index: 296966 (C4 (Web Text))

Max Activating Token Index: 844

Text #11

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.4462. Min Act: -0.1700

Data Index: 374882 (C4 (Web Text))

Max Activating Token Index: 907

Text #12

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.4462. Min Act: -0.1700

Data Index: 593664 (C4 (Web Text))

Max Activating Token Index: 693

Text #13

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.3837. Min Act: -0.1700

Data Index: 67139 (C4 (Web Text))

Max Activating Token Index: 810

Text #14

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.4150. Min Act: -0.1700

Data Index: 427106 (C4 (Web Text))

Max Activating Token Index: 27

Text #15

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.4150. Min Act: -0.1700

Data Index: 1221355 (C4 (Web Text))

Max Activating Token Index: 726

Text #16

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.4150. Min Act: -0.1700

Data Index: 1243571 (C4 (Web Text))

Max Activating Token Index: 989

Text #17

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.4150. Min Act: -0.1700

Data Index: 716508 (C4 (Web Text))

Max Activating Token Index: 936

Text #18

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.3212. Min Act: -0.1700

Data Index: 1258786 (C4 (Web Text))

Max Activating Token Index: 119

Text #19

Max Range: 5.8212. Min Range: -5.8212

Max Act: 5.4150. Min Act: -0.1700

Data Index: 707100 (C4 (Web Text))

Max Activating Token Index: 534