INDEX

Explanations

The neuron fires on the single most salient content word in a sentence or clause—i.e. it picks out the main topic or focus word.

after it completes allows

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

 osal

-1.09

 finally

-1.05

ところに

-0.94

 наконец

-0.94

 있으며

-0.92

之所

-0.90

をもら

-0.90

บาย

-0.89

 pasir

-0.88

来到了

-0.88

POSITIVE LOGITS

)}</

1.05

ハワイ

1.04

">)</

0.96

 době

0.96

unno

0.94

病情

0.93

lwz

0.93

Есть

0.91

Сейчас

0.91

Итак

0.90

Activations Density 0.235%