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

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

Neuron 1398 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-4l')

Text #0

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.8386. Min Act: -0.1700

Data Index: 1338213 (C4 (Web Text))

Max Activating Token Index: 1023

Text #1

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.7448. Min Act: -0.1700

Data Index: 1197133 (C4 (Web Text))

Max Activating Token Index: 1023

Text #2

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.7136. Min Act: -0.1700

Data Index: 1194508 (C4 (Web Text))

Max Activating Token Index: 1023

Text #3

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.7136. Min Act: -0.1700

Data Index: 1220238 (C4 (Web Text))

Max Activating Token Index: 1023

Text #4

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.6511. Min Act: -0.1700

Data Index: 1349096 (C4 (Web Text))

Max Activating Token Index: 52

Text #5

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.6198. Min Act: -0.1700

Data Index: 1071072 (C4 (Web Text))

Max Activating Token Index: 1023

Text #6

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.6198. Min Act: -0.1700

Data Index: 1080269 (C4 (Web Text))

Max Activating Token Index: 1023

Text #7

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.5886. Min Act: -0.1700

Data Index: 1031151 (C4 (Web Text))

Max Activating Token Index: 1023

Text #8

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.6198. Min Act: -0.1700

Data Index: 653261 (C4 (Web Text))

Max Activating Token Index: 1023

Text #9

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.5573. Min Act: -0.1700

Data Index: 494382 (C4 (Web Text))

Max Activating Token Index: 1023

Text #10

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.5573. Min Act: -0.1700

Data Index: 409640 (C4 (Web Text))

Max Activating Token Index: 1023

Text #11

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.5261. Min Act: -0.1700

Data Index: 390614 (C4 (Web Text))

Max Activating Token Index: 1023

Text #12

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.5261. Min Act: -0.1700

Data Index: 566244 (C4 (Web Text))

Max Activating Token Index: 1023

Text #13

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.5261. Min Act: -0.1700

Data Index: 803745 (C4 (Web Text))

Max Activating Token Index: 1023

Text #14

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.4636. Min Act: -0.1700

Data Index: 995458 (C4 (Web Text))

Max Activating Token Index: 1023

Text #15

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.4948. Min Act: -0.1700

Data Index: 1234216 (C4 (Web Text))

Max Activating Token Index: 1023

Text #16

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.5261. Min Act: -0.1700

Data Index: 746094 (C4 (Web Text))

Max Activating Token Index: 902

Text #17

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.4948. Min Act: -0.1700

Data Index: 230958 (C4 (Web Text))

Max Activating Token Index: 1023

Text #18

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.4636. Min Act: -0.1700

Data Index: 1053999 (C4 (Web Text))

Max Activating Token Index: 1023

Text #19

Max Range: 5.8386. Min Range: -5.8386

Max Act: 5.4636. Min Act: -0.1700

Data Index: 419254 (C4 (Web Text))

Max Activating Token Index: 1023