INDEX

Explanations

latitude and longitude

This neuron fires most strongly on salient content words—especially distinctive nouns or key technical terms that often appear in titles, headlines, or question subjects.

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

！】

-1.56

aktion

-1.53

 there

-1.44

 ۲۰۱

-1.34

 Moscú

-1.31

 Depois

-1.30

 testGet

-1.30

要说

-1.27

 ?></

-1.23

 października

-1.23

POSITIVE LOGITS

↵↵

1.76

 already

1.55

 usual

1.29

 current

1.27

 these

1.24

 pouze

1.24

 ähnlichen

1.23

这些

1.22

 solchen

1.20

↵

1.20

Activations Density 0.148%