INDEX

Explanations

standalone="no"

The neuron responds to XML/HTML-style attribute assignments (e.g. name="value" pairs inside tags).

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

listFiles

-0.88

浓郁

-0.78

stretchr

-0.74

Greater

-0.72

 higher

-0.71

悌

-0.71

闖

-0.70

قي

-0.70

 through

-0.70

isies

-0.69

POSITIVE LOGITS

Folding

0.84

 Folding

0.84

🏉

0.83

 GOAL

0.82

folding

0.81

 INCREASE

0.81

LBL

0.80

iredo

0.80

\"");

0.80

 fter

0.80

Activations Density 0.001%