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

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

Neuron 360 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: 3.9901. Min Range: -3.9901

Max Act: 3.9901. Min Act: -0.1700

Data Index: 17502 (C4 (Web Text))

Max Activating Token Index: 732

Text #1

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.9588. Min Act: -0.1700

Data Index: 658401 (C4 (Web Text))

Max Activating Token Index: 377

Text #2

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.9276. Min Act: -0.1700

Data Index: 548490 (C4 (Web Text))

Max Activating Token Index: 448

Text #3

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.8963. Min Act: -0.1700

Data Index: 566519 (C4 (Web Text))

Max Activating Token Index: 508

Text #4

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.8494. Min Act: -0.1700

Data Index: 1169376 (C4 (Web Text))

Max Activating Token Index: 257

Text #5

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.8807. Min Act: -0.1700

Data Index: 162002 (C4 (Web Text))

Max Activating Token Index: 482

Text #6

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.8650. Min Act: -0.1700

Data Index: 479593 (C4 (Web Text))

Max Activating Token Index: 378

Text #7

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.8494. Min Act: -0.1700

Data Index: 913144 (C4 (Web Text))

Max Activating Token Index: 308

Text #8

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.8337. Min Act: -0.1700

Data Index: 1344473 (C4 (Web Text))

Max Activating Token Index: 266

Text #9

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.8181. Min Act: -0.1700

Data Index: 1202569 (C4 (Web Text))

Max Activating Token Index: 92

Text #10

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.8025. Min Act: -0.1700

Data Index: 763430 (C4 (Web Text))

Max Activating Token Index: 234

Text #11

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.7868. Min Act: -0.1700

Data Index: 588399 (C4 (Web Text))

Max Activating Token Index: 596

Text #12

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.7712. Min Act: -0.1700

Data Index: 1273062 (C4 (Web Text))

Max Activating Token Index: 632

Text #13

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.8025. Min Act: -0.1700

Data Index: 60153 (C4 (Web Text))

Max Activating Token Index: 174

Text #14

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.7868. Min Act: -0.1700

Data Index: 1306594 (C4 (Web Text))

Max Activating Token Index: 805

Text #15

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.7555. Min Act: -0.1700

Data Index: 225590 (C4 (Web Text))

Max Activating Token Index: 521

Text #16

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.7712. Min Act: -0.1700

Data Index: 286838 (C4 (Web Text))

Max Activating Token Index: 460

Text #17

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.7399. Min Act: -0.1700

Data Index: 385322 (C4 (Web Text))

Max Activating Token Index: 261

Text #18

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.7399. Min Act: -0.1700

Data Index: 1001992 (C4 (Web Text))

Max Activating Token Index: 596

Text #19

Max Range: 3.9901. Min Range: -3.9901

Max Act: 3.7086. Min Act: -0.1700

Data Index: 1077528 (C4 (Web Text))

Max Activating Token Index: 900