INDEX

Explanations

Japanese and English text snippets

The neuron activates on the Japanese genitive/possessive particle “の.”

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

變

-2.25

 και

-2.23

up

-2.22

 szczeg

-2.11

將

-2.09

硖

-2.08

 або

-2.02

轉

-2.00

啟動

-1.98

兩

-1.95

POSITIVE LOGITS

↵↵

2.05

僦

1.98

を使う

1.94

のような

1.93

1.92

</sub>

1.90

 trabajadores

1.90

 glimmer

1.88

).

1.88

を使用

1.88

Activations Density 0.001%