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

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

Neuron 1652 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.4936. Min Act: -0.1700

Data Index: 8437 (C4 (Web Text))

Max Activating Token Index: 851

Text #1

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.4309. Min Act: -0.1700

Data Index: 1150058 (C4 (Web Text))

Max Activating Token Index: 760

Text #2

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.4465. Min Act: -0.1700

Data Index: 1211189 (C4 (Web Text))

Max Activating Token Index: 926

Text #3

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.4465. Min Act: -0.1700

Data Index: 796177 (C4 (Web Text))

Max Activating Token Index: 658

Text #4

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.4152. Min Act: -0.1700

Data Index: 1525676 (Python Code)

Max Activating Token Index: 520

Text #5

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.3995. Min Act: -0.1700

Data Index: 410417 (C4 (Web Text))

Max Activating Token Index: 353

Text #6

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.4152. Min Act: -0.1700

Data Index: 696900 (C4 (Web Text))

Max Activating Token Index: 535

Text #7

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.3995. Min Act: -0.1700

Data Index: 324605 (C4 (Web Text))

Max Activating Token Index: 491

Text #8

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.3838. Min Act: -0.1700

Data Index: 559089 (C4 (Web Text))

Max Activating Token Index: 341

Text #9

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.3838. Min Act: -0.1700

Data Index: 433876 (C4 (Web Text))

Max Activating Token Index: 308

Text #10

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.3524. Min Act: -0.1700

Data Index: 752289 (C4 (Web Text))

Max Activating Token Index: 664

Text #11

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.3367. Min Act: -0.1700

Data Index: 1345729 (C4 (Web Text))

Max Activating Token Index: 993

Text #12

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.3210. Min Act: -0.1700

Data Index: 219949 (C4 (Web Text))

Max Activating Token Index: 471

Text #13

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.3681. Min Act: -0.1700

Data Index: 231748 (C4 (Web Text))

Max Activating Token Index: 75

Text #14

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.3210. Min Act: -0.1700

Data Index: 494621 (C4 (Web Text))

Max Activating Token Index: 164

Text #15

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.2896. Min Act: -0.1700

Data Index: 1655728 (Python Code)

Max Activating Token Index: 321

Text #16

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.3210. Min Act: -0.1700

Data Index: 296263 (C4 (Web Text))

Max Activating Token Index: 444

Text #17

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.2896. Min Act: -0.1700

Data Index: 198907 (C4 (Web Text))

Max Activating Token Index: 304

Text #18

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.3681. Min Act: -0.1700

Data Index: 1214954 (C4 (Web Text))

Max Activating Token Index: 357

Text #19

Max Range: 3.4936. Min Range: -3.4936

Max Act: 3.2739. Min Act: -0.1700

Data Index: 905516 (C4 (Web Text))

Max Activating Token Index: 941