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

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

Neuron 500 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.2098. Min Act: -0.1700

Data Index: 1435129 (Python Code)

Max Activating Token Index: 569

Text #1

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.1783. Min Act: -0.1700

Data Index: 337176 (C4 (Web Text))

Max Activating Token Index: 811

Text #2

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.2413. Min Act: -0.1700

Data Index: 1522703 (Python Code)

Max Activating Token Index: 510

Text #3

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.2098. Min Act: -0.1700

Data Index: 1587407 (Python Code)

Max Activating Token Index: 815

Text #4

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.2098. Min Act: -0.1700

Data Index: 1422149 (Python Code)

Max Activating Token Index: 963

Text #5

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.1783. Min Act: -0.1700

Data Index: 1617770 (Python Code)

Max Activating Token Index: 1023

Text #6

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.1468. Min Act: -0.1700

Data Index: 390676 (C4 (Web Text))

Max Activating Token Index: 108

Text #7

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.0838. Min Act: -0.1700

Data Index: 1048798 (C4 (Web Text))

Max Activating Token Index: 134

Text #8

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.0838. Min Act: -0.1700

Data Index: 1370794 (Python Code)

Max Activating Token Index: 709

Text #9

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.0838. Min Act: -0.1700

Data Index: 1508218 (Python Code)

Max Activating Token Index: 969

Text #10

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.0838. Min Act: -0.1700

Data Index: 1401915 (Python Code)

Max Activating Token Index: 285

Text #11

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.0838. Min Act: -0.1700

Data Index: 942854 (C4 (Web Text))

Max Activating Token Index: 69

Text #12

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.0522. Min Act: -0.1699

Data Index: 989594 (C4 (Web Text))

Max Activating Token Index: 132

Text #13

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.0838. Min Act: -0.1700

Data Index: 1560413 (Python Code)

Max Activating Token Index: 942

Text #14

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.0522. Min Act: -0.1700

Data Index: 1629154 (Python Code)

Max Activating Token Index: 819

Text #15

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.0207. Min Act: -0.1700

Data Index: 1623638 (Python Code)

Max Activating Token Index: 625

Text #16

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.0522. Min Act: -0.1700

Data Index: 1430076 (Python Code)

Max Activating Token Index: 144

Text #17

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.0207. Min Act: -0.1700

Data Index: 973738 (C4 (Web Text))

Max Activating Token Index: 96

Text #18

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.0207. Min Act: -0.1700

Data Index: 1556604 (Python Code)

Max Activating Token Index: 637

Text #19

Max Range: 3.2413. Min Range: -3.2413

Max Act: 3.0207. Min Act: -0.1700

Data Index: 1430722 (Python Code)

Max Activating Token Index: 788