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

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

Neuron 1959 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-3l')

Text #0

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.5103. Min Act: -0.1700

Data Index: 1025812 (C4 (Web Text))

Max Activating Token Index: 697

Text #1

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.4478. Min Act: -0.1700

Data Index: 969916 (C4 (Web Text))

Max Activating Token Index: 556

Text #2

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.4478. Min Act: -0.1700

Data Index: 316473 (C4 (Web Text))

Max Activating Token Index: 657

Text #3

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.1978. Min Act: -0.1700

Data Index: 60802 (C4 (Web Text))

Max Activating Token Index: 924

Text #4

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.2603. Min Act: -0.1700

Data Index: 252843 (C4 (Web Text))

Max Activating Token Index: 112

Text #5

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.0728. Min Act: -0.1700

Data Index: 1119739 (C4 (Web Text))

Max Activating Token Index: 679

Text #6

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.2603. Min Act: -0.1700

Data Index: 1190778 (C4 (Web Text))

Max Activating Token Index: 549

Text #7

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.2603. Min Act: -0.1700

Data Index: 1282206 (C4 (Web Text))

Max Activating Token Index: 986

Text #8

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.1978. Min Act: -0.1700

Data Index: 1109230 (C4 (Web Text))

Max Activating Token Index: 24

Text #9

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.1353. Min Act: -0.1700

Data Index: 247427 (C4 (Web Text))

Max Activating Token Index: 18

Text #10

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.1353. Min Act: -0.1700

Data Index: 395670 (C4 (Web Text))

Max Activating Token Index: 811

Text #11

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.1353. Min Act: -0.1700

Data Index: 772142 (C4 (Web Text))

Max Activating Token Index: 386

Text #12

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.0103. Min Act: -0.1700

Data Index: 1096004 (C4 (Web Text))

Max Activating Token Index: 701

Text #13

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.0728. Min Act: -0.1700

Data Index: 1224550 (C4 (Web Text))

Max Activating Token Index: 635

Text #14

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.0728. Min Act: -0.1700

Data Index: 666414 (C4 (Web Text))

Max Activating Token Index: 344

Text #15

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.0103. Min Act: -0.1700

Data Index: 27812 (C4 (Web Text))

Max Activating Token Index: 541

Text #16

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.0728. Min Act: -0.1699

Data Index: 978564 (C4 (Web Text))

Max Activating Token Index: 981

Text #17

Max Range: 10.5103. Min Range: -10.5103

Max Act: 10.0103. Min Act: -0.1700

Data Index: 186218 (C4 (Web Text))

Max Activating Token Index: 720

Text #18

Max Range: 10.5103. Min Range: -10.5103

Max Act: 9.9478. Min Act: -0.1700

Data Index: 607040 (C4 (Web Text))

Max Activating Token Index: 876

Text #19

Max Range: 10.5103. Min Range: -10.5103

Max Act: 9.8853. Min Act: -0.1699

Data Index: 574784 (C4 (Web Text))

Max Activating Token Index: 698