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

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

Neuron 970 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: 5.1983. Min Range: -5.1983

Max Act: 5.1983. Min Act: -0.1700

Data Index: 87824 (C4 (Web Text))

Max Activating Token Index: 234

Text #1

Max Range: 5.1983. Min Range: -5.1983

Max Act: 5.0733. Min Act: -0.1700

Data Index: 298112 (C4 (Web Text))

Max Activating Token Index: 9

Text #2

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.9796. Min Act: -0.1700

Data Index: 386943 (C4 (Web Text))

Max Activating Token Index: 786

Text #3

Max Range: 5.1983. Min Range: -5.1983

Max Act: 5.0108. Min Act: -0.1700

Data Index: 1284031 (C4 (Web Text))

Max Activating Token Index: 10

Text #4

Max Range: 5.1983. Min Range: -5.1983

Max Act: 5.0733. Min Act: -0.1700

Data Index: 1142556 (C4 (Web Text))

Max Activating Token Index: 882

Text #5

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.9796. Min Act: -0.1700

Data Index: 665783 (C4 (Web Text))

Max Activating Token Index: 640

Text #6

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.9483. Min Act: -0.1700

Data Index: 1295478 (C4 (Web Text))

Max Activating Token Index: 541

Text #7

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.9483. Min Act: -0.1700

Data Index: 898201 (C4 (Web Text))

Max Activating Token Index: 92

Text #8

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.9171. Min Act: -0.1700

Data Index: 185859 (C4 (Web Text))

Max Activating Token Index: 648

Text #9

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.9483. Min Act: -0.1700

Data Index: 1086448 (C4 (Web Text))

Max Activating Token Index: 508

Text #10

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.9171. Min Act: -0.1700

Data Index: 190522 (C4 (Web Text))

Max Activating Token Index: 323

Text #11

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.8858. Min Act: -0.1700

Data Index: 186196 (C4 (Web Text))

Max Activating Token Index: 392

Text #12

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.8233. Min Act: -0.1700

Data Index: 397407 (C4 (Web Text))

Max Activating Token Index: 698

Text #13

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.8546. Min Act: -0.1700

Data Index: 52478 (C4 (Web Text))

Max Activating Token Index: 29

Text #14

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.8233. Min Act: -0.1700

Data Index: 918439 (C4 (Web Text))

Max Activating Token Index: 885

Text #15

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.8233. Min Act: -0.1700

Data Index: 652722 (C4 (Web Text))

Max Activating Token Index: 440

Text #16

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.8546. Min Act: -0.1699

Data Index: 1242022 (C4 (Web Text))

Max Activating Token Index: 383

Text #17

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.8546. Min Act: -0.1700

Data Index: 302137 (C4 (Web Text))

Max Activating Token Index: 541

Text #18

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.8233. Min Act: -0.1700

Data Index: 545739 (C4 (Web Text))

Max Activating Token Index: 612

Text #19

Max Range: 5.1983. Min Range: -5.1983

Max Act: 4.8546. Min Act: -0.1700

Data Index: 1024910 (C4 (Web Text))

Max Activating Token Index: 160