INDEX

Explanations

originally previously

This neuron responds to numerical and financial‐reporting tokens—amounts, percentages, dates, share‐price movements, and other figures common in business news.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 veramente

0.77

 prostu

0.71

ちゃんと

0.70

0.67

0.66

 সত্যিকার

0.66

 devemos

0.66

we

0.64

 그냥

0.64

の中に

0.64

POSITIVE LOGITS

 지난해

1.38

此前

1.30

 reportedly

1.11

 atualmente

1.05

去年

1.02

据悉

1.02

 ఇటీవల

1.02

近年来

0.96

 फिलहाल

0.94

 ప్రస్తుతం

0.93

Activations Density 1.167%