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

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

Neuron 1004 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: 5.4520. Min Range: -5.4520

Max Act: 5.4520. Min Act: -0.1700

Data Index: 566780 (C4 (Web Text))

Max Activating Token Index: 566

Text #1

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.2958. Min Act: -0.1700

Data Index: 1178865 (C4 (Web Text))

Max Activating Token Index: 657

Text #2

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.2645. Min Act: -0.1700

Data Index: 315013 (C4 (Web Text))

Max Activating Token Index: 360

Text #3

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.2020. Min Act: -0.1700

Data Index: 402834 (C4 (Web Text))

Max Activating Token Index: 342

Text #4

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.1083. Min Act: -0.1700

Data Index: 681268 (C4 (Web Text))

Max Activating Token Index: 958

Text #5

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.1083. Min Act: -0.1700

Data Index: 552947 (C4 (Web Text))

Max Activating Token Index: 782

Text #6

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.2020. Min Act: -0.1700

Data Index: 37899 (C4 (Web Text))

Max Activating Token Index: 264

Text #7

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.1708. Min Act: -0.1700

Data Index: 834565 (C4 (Web Text))

Max Activating Token Index: 711

Text #8

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.1395. Min Act: -0.1700

Data Index: 1075098 (C4 (Web Text))

Max Activating Token Index: 144

Text #9

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.1083. Min Act: -0.1700

Data Index: 427858 (C4 (Web Text))

Max Activating Token Index: 682

Text #10

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.1083. Min Act: -0.1700

Data Index: 322279 (C4 (Web Text))

Max Activating Token Index: 687

Text #11

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.1083. Min Act: -0.1700

Data Index: 1185133 (C4 (Web Text))

Max Activating Token Index: 926

Text #12

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.0770. Min Act: -0.1700

Data Index: 298101 (C4 (Web Text))

Max Activating Token Index: 731

Text #13

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.0145. Min Act: -0.1700

Data Index: 130011 (C4 (Web Text))

Max Activating Token Index: 904

Text #14

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.0458. Min Act: -0.1700

Data Index: 495704 (C4 (Web Text))

Max Activating Token Index: 593

Text #15

Max Range: 5.4520. Min Range: -5.4520

Max Act: 4.9520. Min Act: -0.1700

Data Index: 894949 (C4 (Web Text))

Max Activating Token Index: 835

Text #16

Max Range: 5.4520. Min Range: -5.4520

Max Act: 5.0458. Min Act: -0.1700

Data Index: 1031037 (C4 (Web Text))

Max Activating Token Index: 340

Text #17

Max Range: 5.4520. Min Range: -5.4520

Max Act: 4.9520. Min Act: -0.1700

Data Index: 690346 (C4 (Web Text))

Max Activating Token Index: 443

Text #18

Max Range: 5.4520. Min Range: -5.4520

Max Act: 4.9833. Min Act: -0.1700

Data Index: 632013 (C4 (Web Text))

Max Activating Token Index: 1017

Text #19

Max Range: 5.4520. Min Range: -5.4520

Max Act: 4.9833. Min Act: -0.1700

Data Index: 1111817 (C4 (Web Text))

Max Activating Token Index: 378