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

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

Neuron 478 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: 3.3546. Min Range: -3.3546

Max Act: 3.3546. Min Act: -0.1700

Data Index: 524791 (C4 (Web Text))

Max Activating Token Index: 828

Text #1

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.3546. Min Act: -0.1700

Data Index: 246718 (C4 (Web Text))

Max Activating Token Index: 1023

Text #2

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.2604. Min Act: -0.1700

Data Index: 579525 (C4 (Web Text))

Max Activating Token Index: 1023

Text #3

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.2132. Min Act: -0.1700

Data Index: 1494539 (Python Code)

Max Activating Token Index: 1023

Text #4

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.1817. Min Act: -0.1700

Data Index: 936049 (C4 (Web Text))

Max Activating Token Index: 250

Text #5

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.1503. Min Act: -0.1700

Data Index: 468681 (C4 (Web Text))

Max Activating Token Index: 401

Text #6

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.1660. Min Act: -0.1700

Data Index: 41490 (C4 (Web Text))

Max Activating Token Index: 38

Text #7

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.1345. Min Act: -0.1700

Data Index: 4312 (C4 (Web Text))

Max Activating Token Index: 1023

Text #8

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.1345. Min Act: -0.1700

Data Index: 1160866 (C4 (Web Text))

Max Activating Token Index: 474

Text #9

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.1030. Min Act: -0.1700

Data Index: 1212226 (C4 (Web Text))

Max Activating Token Index: 1023

Text #10

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.1030. Min Act: -0.1700

Data Index: 1551865 (Python Code)

Max Activating Token Index: 1023

Text #11

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.0872. Min Act: -0.1700

Data Index: 1047529 (C4 (Web Text))

Max Activating Token Index: 1023

Text #12

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.1030. Min Act: -0.1700

Data Index: 93059 (C4 (Web Text))

Max Activating Token Index: 1023

Text #13

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.0557. Min Act: -0.1700

Data Index: 15532 (C4 (Web Text))

Max Activating Token Index: 724

Text #14

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.0399. Min Act: -0.1700

Data Index: 977476 (C4 (Web Text))

Max Activating Token Index: 923

Text #15

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.0241. Min Act: -0.1700

Data Index: 167983 (C4 (Web Text))

Max Activating Token Index: 107

Text #16

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.0083. Min Act: -0.1700

Data Index: 690230 (C4 (Web Text))

Max Activating Token Index: 568

Text #17

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.0083. Min Act: -0.1700

Data Index: 333040 (C4 (Web Text))

Max Activating Token Index: 662

Text #18

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.0241. Min Act: -0.1700

Data Index: 371722 (C4 (Web Text))

Max Activating Token Index: 329

Text #19

Max Range: 3.3546. Min Range: -3.3546

Max Act: 3.0241. Min Act: -0.1700

Data Index: 646303 (C4 (Web Text))

Max Activating Token Index: 162