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

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

Neuron 2014 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: 3.9061. Min Range: -3.9061

Max Act: 3.9061. Min Act: -0.1700

Data Index: 267709 (C4 (Web Text))

Max Activating Token Index: 933

Text #1

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.9061. Min Act: -0.1700

Data Index: 1165335 (C4 (Web Text))

Max Activating Token Index: 303

Text #2

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8749. Min Act: -0.1700

Data Index: 1171110 (C4 (Web Text))

Max Activating Token Index: 858

Text #3

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8749. Min Act: -0.1700

Data Index: 807221 (C4 (Web Text))

Max Activating Token Index: 601

Text #4

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8749. Min Act: -0.1700

Data Index: 484083 (C4 (Web Text))

Max Activating Token Index: 818

Text #5

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8123. Min Act: -0.1700

Data Index: 338321 (C4 (Web Text))

Max Activating Token Index: 904

Text #6

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8436. Min Act: -0.1700

Data Index: 1151699 (C4 (Web Text))

Max Activating Token Index: 145

Text #7

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8436. Min Act: -0.1700

Data Index: 571040 (C4 (Web Text))

Max Activating Token Index: 701

Text #8

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8123. Min Act: -0.1700

Data Index: 1094014 (C4 (Web Text))

Max Activating Token Index: 878

Text #9

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8123. Min Act: -0.1700

Data Index: 1012693 (C4 (Web Text))

Max Activating Token Index: 298

Text #10

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8123. Min Act: -0.1700

Data Index: 1195026 (C4 (Web Text))

Max Activating Token Index: 779

Text #11

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8123. Min Act: -0.1700

Data Index: 576649 (C4 (Web Text))

Max Activating Token Index: 974

Text #12

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.7810. Min Act: -0.1700

Data Index: 1313939 (C4 (Web Text))

Max Activating Token Index: 1023

Text #13

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.7810. Min Act: -0.1700

Data Index: 1204782 (C4 (Web Text))

Max Activating Token Index: 874

Text #14

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8123. Min Act: -0.1700

Data Index: 145246 (C4 (Web Text))

Max Activating Token Index: 764

Text #15

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8123. Min Act: -0.1700

Data Index: 361373 (C4 (Web Text))

Max Activating Token Index: 467

Text #16

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8123. Min Act: -0.1700

Data Index: 861906 (C4 (Web Text))

Max Activating Token Index: 353

Text #17

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.7810. Min Act: -0.1700

Data Index: 275822 (C4 (Web Text))

Max Activating Token Index: 15

Text #18

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8123. Min Act: -0.1700

Data Index: 339090 (C4 (Web Text))

Max Activating Token Index: 341

Text #19

Max Range: 3.9061. Min Range: -3.9061

Max Act: 3.8123. Min Act: -0.1700

Data Index: 1147251 (C4 (Web Text))

Max Activating Token Index: 715