INDEX

Explanations

specific tools and guides

The neuron fires on nouns naming features, tools, or functional elements (e.g. “timer,” “tutorial,” “border,” “tour,” “newsletter,” “summary,” etc.).

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

的就

-1.45

 entirety

-1.22

without

-1.19

給我

-1.17

-1.16

 Zweck

-1.15

喏

-1.15

 lacking

-1.13

-1.11

POSITIVE LOGITS

 that

2.44

for

2.38

of

2.05

to

1.78

iation

1.70

[]

1.64

 hanggang

1.62

sssss

1.61

 faſt

1.61

ssssssss

1.59

Activations Density 0.082%