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

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

Neuron 942 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: 7.7847. Min Range: -7.7847

Max Act: 7.7847. Min Act: -0.1700

Data Index: 1035576 (C4 (Web Text))

Max Activating Token Index: 629

Text #1

Max Range: 7.7847. Min Range: -7.7847

Max Act: 7.7535. Min Act: -0.1700

Data Index: 291413 (C4 (Web Text))

Max Activating Token Index: 898

Text #2

Max Range: 7.7847. Min Range: -7.7847

Max Act: 7.5972. Min Act: -0.1700

Data Index: 1089966 (C4 (Web Text))

Max Activating Token Index: 494

Text #3

Max Range: 7.7847. Min Range: -7.7847

Max Act: 7.4722. Min Act: -0.1700

Data Index: 220176 (C4 (Web Text))

Max Activating Token Index: 829

Text #4

Max Range: 7.7847. Min Range: -7.7847

Max Act: 7.4410. Min Act: -0.1700

Data Index: 1097834 (C4 (Web Text))

Max Activating Token Index: 705

Text #5

Max Range: 7.7847. Min Range: -7.7847

Max Act: 7.3785. Min Act: -0.1700

Data Index: 1160241 (C4 (Web Text))

Max Activating Token Index: 936

Text #6

Max Range: 7.7847. Min Range: -7.7847

Max Act: 7.1597. Min Act: -0.1700

Data Index: 699733 (C4 (Web Text))

Max Activating Token Index: 430

Text #7

Max Range: 7.7847. Min Range: -7.7847

Max Act: 7.2222. Min Act: -0.1699

Data Index: 1514527 (Python Code)

Max Activating Token Index: 987

Text #8

Max Range: 7.7847. Min Range: -7.7847

Max Act: 7.1910. Min Act: -0.1700

Data Index: 761876 (C4 (Web Text))

Max Activating Token Index: 757

Text #9

Max Range: 7.7847. Min Range: -7.7847

Max Act: 7.0347. Min Act: -0.1700

Data Index: 70267 (C4 (Web Text))

Max Activating Token Index: 698

Text #10

Max Range: 7.7847. Min Range: -7.7847

Max Act: 7.0347. Min Act: -0.1700

Data Index: 1304992 (C4 (Web Text))

Max Activating Token Index: 261

Text #11

Max Range: 7.7847. Min Range: -7.7847

Max Act: 7.0347. Min Act: -0.1699

Data Index: 836351 (C4 (Web Text))

Max Activating Token Index: 859

Text #12

Max Range: 7.7847. Min Range: -7.7847

Max Act: 7.0347. Min Act: -0.1700

Data Index: 562977 (C4 (Web Text))

Max Activating Token Index: 222

Text #13

Max Range: 7.7847. Min Range: -7.7847

Max Act: 7.0035. Min Act: -0.1700

Data Index: 1000734 (C4 (Web Text))

Max Activating Token Index: 276

Text #14

Max Range: 7.7847. Min Range: -7.7847

Max Act: 6.9097. Min Act: -0.1700

Data Index: 1609098 (Python Code)

Max Activating Token Index: 130

Text #15

Max Range: 7.7847. Min Range: -7.7847

Max Act: 6.9722. Min Act: -0.1700

Data Index: 1552998 (Python Code)

Max Activating Token Index: 414

Text #16

Max Range: 7.7847. Min Range: -7.7847

Max Act: 6.8785. Min Act: -0.1700

Data Index: 643640 (C4 (Web Text))

Max Activating Token Index: 545

Text #17

Max Range: 7.7847. Min Range: -7.7847

Max Act: 6.8785. Min Act: -0.1700

Data Index: 1535038 (Python Code)

Max Activating Token Index: 577

Text #18

Max Range: 7.7847. Min Range: -7.7847

Max Act: 6.8472. Min Act: -0.1700

Data Index: 1499112 (Python Code)

Max Activating Token Index: 886

Text #19

Max Range: 7.7847. Min Range: -7.7847

Max Act: 6.9097. Min Act: -0.1700

Data Index: 1591457 (Python Code)

Max Activating Token Index: 124