INDEX

Explanations

student grade level

The neuron activates on terms that refer to academic status and schooling context—e.g. class standing (freshman, junior, senior), grades, majors, and related education‐level markers.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

by

-1.31

 both

-1.29

 INSTR

-1.20

 Both

-1.16

for

-1.13

tū

-1.06

客様

-1.05

！（

-1.05

 many

-1.04

POSITIVE LOGITS

ৢ

1.60

who

1.50

问我

1.48

 EVEN

1.42

頑張って

1.38

 fluxo

1.38

 prioridades

1.34

cault

1.32

 особенно

1.30

淠

1.30

Activations Density 0.038%