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

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

Neuron 1617 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: 3.9579. Min Range: -3.9579

Max Act: 3.9579. Min Act: -0.1700

Data Index: 448377 (C4 (Web Text))

Max Activating Token Index: 887

Text #1

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.8172. Min Act: -0.1700

Data Index: 1299650 (C4 (Web Text))

Max Activating Token Index: 462

Text #2

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.8328. Min Act: -0.1700

Data Index: 259610 (C4 (Web Text))

Max Activating Token Index: 571

Text #3

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.8015. Min Act: -0.1700

Data Index: 998447 (C4 (Web Text))

Max Activating Token Index: 415

Text #4

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.7859. Min Act: -0.1700

Data Index: 859831 (C4 (Web Text))

Max Activating Token Index: 390

Text #5

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.7702. Min Act: -0.1700

Data Index: 79017 (C4 (Web Text))

Max Activating Token Index: 440

Text #6

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.7546. Min Act: -0.1700

Data Index: 997652 (C4 (Web Text))

Max Activating Token Index: 617

Text #7

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.7546. Min Act: -0.1700

Data Index: 1179659 (C4 (Web Text))

Max Activating Token Index: 802

Text #8

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.7546. Min Act: -0.1700

Data Index: 520187 (C4 (Web Text))

Max Activating Token Index: 263

Text #9

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.7390. Min Act: -0.1700

Data Index: 1338921 (C4 (Web Text))

Max Activating Token Index: 897

Text #10

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.7077. Min Act: -0.1700

Data Index: 1068061 (C4 (Web Text))

Max Activating Token Index: 523

Text #11

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.7077. Min Act: -0.1700

Data Index: 21745 (C4 (Web Text))

Max Activating Token Index: 241

Text #12

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.6764. Min Act: -0.1700

Data Index: 724059 (C4 (Web Text))

Max Activating Token Index: 542

Text #13

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.6607. Min Act: -0.1700

Data Index: 1299661 (C4 (Web Text))

Max Activating Token Index: 378

Text #14

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.6764. Min Act: -0.1700

Data Index: 829877 (C4 (Web Text))

Max Activating Token Index: 727

Text #15

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.6920. Min Act: -0.1700

Data Index: 938593 (C4 (Web Text))

Max Activating Token Index: 926

Text #16

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.6451. Min Act: -0.1700

Data Index: 509176 (C4 (Web Text))

Max Activating Token Index: 588

Text #17

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.6764. Min Act: -0.1700

Data Index: 238112 (C4 (Web Text))

Max Activating Token Index: 910

Text #18

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.6607. Min Act: -0.1700

Data Index: 247238 (C4 (Web Text))

Max Activating Token Index: 753

Text #19

Max Range: 3.9579. Min Range: -3.9579

Max Act: 3.6607. Min Act: -0.1700

Data Index: 393531 (C4 (Web Text))

Max Activating Token Index: 996