INDEX

Explanations

beta-carotene

This neuron activates for mentions of "carrot" and related terms like "carotene" in scientific or culinary contexts.

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

 Sopho

-0.88

Dart

-0.83

衩

-0.83

蚣

-0.79

 autorización

-0.76

 Peppermint

-0.75

 toul

-0.75

 Hibiscus

-0.74

 Donau

-0.73

 miguel

-0.73

POSITIVE LOGITS

 carrot

2.97

 carrots

2.91

🥕

2.47

carrot

2.34

Carrot

2.17

 Carrot

2.13

 Carrots

1.98

 carro

1.59

 carottes

1.38

carro

1.33

Activations Density 0.022%