Exploring Convolutional Neural Networks for Rice Grain Classification: An Explainable AI Approach

Key White House officials are urging lawmakers on Capitol Hill to exclude AI chip export restrictions to China from the annual defense policy bill. This push reflects ongoing strategic considerations regarding technology and national security as the U.S. navigates its relationship with China.

US President Donald Trump stated that investment in artificial intelligence (AI) is significantly benefiting the economy. However, he expressed concerns that overregulation could hinder the US's competitive edge against China in the AI sector.

US officials are reportedly indicating that semiconductor tariffs may not be imposed soon, reflecting a more cautious stance from the Trump administration regarding China. This potential delay comes amid ongoing trade tensions and strategic considerations in US-China relations.

Experts warn that reliance on foreign-made open models poses significant risks to the US supply chain and innovation in artificial intelligence. This dependency could hinder the country's competitive edge against China in the rapidly evolving tech landscape.

China has urged the United States to refrain from interfering in Venezuela's internal affairs, in response to the U.S.'s increasing military presence in the Caribbean and Eastern Pacific. This statement reflects China's position on maintaining peace in Latin America.

The Netherlands has temporarily halted its efforts to seize control of the Chinese-owned chipmaker Nexperia, a move seen as a gesture of goodwill aimed at easing tensions in the ongoing EU-China dispute. This decision comes amid concerns over the impact on global car supply chains, which have been affected by semiconductor shortages.

A new mega data center is set to be constructed in eastern Germany, aimed at establishing a foundation for a competitive European AI sector against the United States and China. This initiative is part of Germany's broader strategy to achieve digital sovereignty.

The paper introduces ScoresActivation, a new activation function designed to enhance global explainability in deep learning models. Traditional post hoc explanation methods often lack connection to the model training process, which can compromise their reliability. ScoresActivation integrates feature importance estimation directly into the training pipeline, allowing models to prioritize features based on their predictive contributions. Evaluations demonstrate that this method produces stable feature rankings that align with established metrics like SHAP values.