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

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

Neuron 101 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: 3.6701. Min Range: -3.6701

Max Act: 3.6388. Min Act: -0.1700

Data Index: 602343 (C4 (Web Text))

Max Activating Token Index: 163

Text #1

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.6701. Min Act: -0.1700

Data Index: 380992 (C4 (Web Text))

Max Activating Token Index: 166

Text #2

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.6075. Min Act: -0.1700

Data Index: 894041 (C4 (Web Text))

Max Activating Token Index: 558

Text #3

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.6075. Min Act: -0.1700

Data Index: 558961 (C4 (Web Text))

Max Activating Token Index: 215

Text #4

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.5448. Min Act: -0.1700

Data Index: 267497 (C4 (Web Text))

Max Activating Token Index: 542

Text #5

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.5448. Min Act: -0.1700

Data Index: 960771 (C4 (Web Text))

Max Activating Token Index: 713

Text #6

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.5135. Min Act: -0.1700

Data Index: 423894 (C4 (Web Text))

Max Activating Token Index: 443

Text #7

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.4822. Min Act: -0.1700

Data Index: 287712 (C4 (Web Text))

Max Activating Token Index: 906

Text #8

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.4822. Min Act: -0.1700

Data Index: 1133297 (C4 (Web Text))

Max Activating Token Index: 605

Text #9

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.4508. Min Act: -0.1700

Data Index: 319774 (C4 (Web Text))

Max Activating Token Index: 974

Text #10

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.4508. Min Act: -0.1700

Data Index: 1064626 (C4 (Web Text))

Max Activating Token Index: 101

Text #11

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.4508. Min Act: -0.1700

Data Index: 1135092 (C4 (Web Text))

Max Activating Token Index: 152

Text #12

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.3881. Min Act: -0.1700

Data Index: 400903 (C4 (Web Text))

Max Activating Token Index: 839

Text #13

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.4822. Min Act: -0.1700

Data Index: 565493 (C4 (Web Text))

Max Activating Token Index: 788

Text #14

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.4508. Min Act: -0.1700

Data Index: 633697 (C4 (Web Text))

Max Activating Token Index: 331

Text #15

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.4508. Min Act: -0.1700

Data Index: 903052 (C4 (Web Text))

Max Activating Token Index: 200

Text #16

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.4195. Min Act: -0.1700

Data Index: 637303 (C4 (Web Text))

Max Activating Token Index: 701

Text #17

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.4195. Min Act: -0.1700

Data Index: 369015 (C4 (Web Text))

Max Activating Token Index: 249

Text #18

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.4508. Min Act: -0.1700

Data Index: 763431 (C4 (Web Text))

Max Activating Token Index: 553

Text #19

Max Range: 3.6701. Min Range: -3.6701

Max Act: 3.3881. Min Act: -0.1700

Data Index: 844899 (C4 (Web Text))

Max Activating Token Index: 742