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

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

Neuron 1627 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.4753. Min Act: -0.1700

Data Index: 1638034 (Python Code)

Max Activating Token Index: 643

Text #1

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.4439. Min Act: -0.1700

Data Index: 1582456 (Python Code)

Max Activating Token Index: 783

Text #2

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.4439. Min Act: -0.1698

Data Index: 1450974 (Python Code)

Max Activating Token Index: 1004

Text #3

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.4753. Min Act: -0.1700

Data Index: 1496231 (Python Code)

Max Activating Token Index: 346

Text #4

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.3812. Min Act: -0.1700

Data Index: 1379735 (Python Code)

Max Activating Token Index: 1009

Text #5

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.3812. Min Act: -0.1700

Data Index: 1404597 (Python Code)

Max Activating Token Index: 531

Text #6

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.3655. Min Act: -0.1700

Data Index: 1498303 (Python Code)

Max Activating Token Index: 155

Text #7

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.3498. Min Act: -0.1700

Data Index: 1567816 (Python Code)

Max Activating Token Index: 5

Text #8

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.3498. Min Act: -0.1700

Data Index: 1558894 (Python Code)

Max Activating Token Index: 669

Text #9

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.3341. Min Act: -0.1700

Data Index: 1431153 (Python Code)

Max Activating Token Index: 644

Text #10

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.2870. Min Act: -0.1700

Data Index: 1429322 (Python Code)

Max Activating Token Index: 267

Text #11

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.3184. Min Act: -0.1700

Data Index: 1578729 (Python Code)

Max Activating Token Index: 379

Text #12

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.2870. Min Act: -0.1700

Data Index: 1455723 (Python Code)

Max Activating Token Index: 921

Text #13

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.3184. Min Act: -0.1700

Data Index: 1476073 (Python Code)

Max Activating Token Index: 923

Text #14

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.3027. Min Act: -0.1700

Data Index: 1432504 (Python Code)

Max Activating Token Index: 716

Text #15

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.2870. Min Act: -0.1700

Data Index: 1500668 (Python Code)

Max Activating Token Index: 380

Text #16

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.3184. Min Act: -0.1700

Data Index: 1303559 (C4 (Web Text))

Max Activating Token Index: 71

Text #17

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.3027. Min Act: -0.1700

Data Index: 1578249 (Python Code)

Max Activating Token Index: 632

Text #18

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.2870. Min Act: -0.1700

Data Index: 1420988 (Python Code)

Max Activating Token Index: 670

Text #19

Max Range: 3.4753. Min Range: -3.4753

Max Act: 3.2713. Min Act: -0.1700

Data Index: 1493152 (Python Code)

Max Activating Token Index: 88