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

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

Neuron 83 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: 3.8374. Min Range: -3.8374

Max Act: 3.8061. Min Act: -0.1700

Data Index: 772700 (C4 (Web Text))

Max Activating Token Index: 412

Text #1

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.8374. Min Act: -0.1700

Data Index: 392286 (C4 (Web Text))

Max Activating Token Index: 863

Text #2

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.8061. Min Act: -0.1700

Data Index: 347754 (C4 (Web Text))

Max Activating Token Index: 981

Text #3

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.7904. Min Act: -0.1700

Data Index: 132151 (C4 (Web Text))

Max Activating Token Index: 580

Text #4

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.7122. Min Act: -0.1700

Data Index: 926428 (C4 (Web Text))

Max Activating Token Index: 857

Text #5

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.7591. Min Act: -0.1700

Data Index: 1341519 (C4 (Web Text))

Max Activating Token Index: 796

Text #6

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.7748. Min Act: -0.1700

Data Index: 962050 (C4 (Web Text))

Max Activating Token Index: 387

Text #7

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.7122. Min Act: -0.1700

Data Index: 676555 (C4 (Web Text))

Max Activating Token Index: 847

Text #8

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.6653. Min Act: -0.1700

Data Index: 851432 (C4 (Web Text))

Max Activating Token Index: 707

Text #9

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.6496. Min Act: -0.1700

Data Index: 1300317 (C4 (Web Text))

Max Activating Token Index: 283

Text #10

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.6340. Min Act: -0.1700

Data Index: 544674 (C4 (Web Text))

Max Activating Token Index: 258

Text #11

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.6809. Min Act: -0.1700

Data Index: 869218 (C4 (Web Text))

Max Activating Token Index: 908

Text #12

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.6966. Min Act: -0.1700

Data Index: 1312844 (C4 (Web Text))

Max Activating Token Index: 432

Text #13

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.6496. Min Act: -0.1700

Data Index: 866585 (C4 (Web Text))

Max Activating Token Index: 910

Text #14

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.6496. Min Act: -0.1700

Data Index: 287233 (C4 (Web Text))

Max Activating Token Index: 989

Text #15

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.5870. Min Act: -0.1700

Data Index: 921290 (C4 (Web Text))

Max Activating Token Index: 732

Text #16

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.6183. Min Act: -0.1700

Data Index: 247905 (C4 (Web Text))

Max Activating Token Index: 322

Text #17

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.5870. Min Act: -0.1700

Data Index: 540236 (C4 (Web Text))

Max Activating Token Index: 201

Text #18

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.5243. Min Act: -0.1700

Data Index: 1088778 (C4 (Web Text))

Max Activating Token Index: 891

Text #19

Max Range: 3.8374. Min Range: -3.8374

Max Act: 3.5713. Min Act: -0.1700

Data Index: 579052 (C4 (Web Text))

Max Activating Token Index: 790