INDEX

Explanations

Official documents and designations

The neuron strongly activates on occurrences of the word “Official.”

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

蠖

-2.72

-2.59

↵↵

-2.58

 hinder

-2.45

-2.44

we

-2.42

 jeśli

-2.41

 architekt

-2.39

 серебря

-2.36

 capítulos

-2.33

POSITIVE LOGITS

?”

3.38

!”

3.34

sii

3.13

酲

3.13

—“

3.05

猹

3.03

椟

3.02

”:

2.98

 stuks

2.98

⇰

2.97

Activations Density 0.014%