INDEX

Explanations

dart related terms

The neuron fires on any token containing the substring “dart,” i.e. it detects occurrences of “dart” (in names, module imports, 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

柵

-2.89

 erectile

-2.78

 muñecas

-2.72

手绘

-2.69

 หรือ

-2.67

聍

-2.56

 всѣ

-2.44

 hijas

-2.41

 обороны

-2.36

 plongée

-2.34

POSITIVE LOGITS

↵↵

2.58

2.45

2.34

--

2.31

 focuses

2.30

2.23

汌

2.11

 realizando

2.09

 termed

2.08

Activations Density 0.002%