Meta-Computing Enhanced Federated Learning in IIoT: Satisfaction-Aware Incentive Scheme via DRL-Based Stackelberg Game

A coalition of state attorneys general has issued a warning to major tech companies, including Meta, Microsoft, Google, and Apple, regarding the potential legal implications of their chatbots producing what they describe as 'delusional outputs.' This letter emphasizes concerns that such outputs may violate state laws, highlighting the need for accountability in AI technologies.

Meta is reportedly developing a new AI model named 'Avocado,' which may not be open source. This initiative reflects the company's ongoing commitment to advancing its artificial intelligence capabilities amidst a competitive landscape.

Meta is reportedly transitioning from its open Llama models to a new AI model codenamed 'Avocado,' which is expected to be a closed model aimed at direct sales, with a potential release in spring 2026. This shift indicates a strategic pivot in Meta's approach to AI development.

Meta's leading artificial intelligence team is experiencing internal friction, creating an us-versus-them mentality with long-standing associates of CEO Mark Zuckerberg. This discord arises amid significant changes in the company's AI initiatives and strategic direction.

Google’s Tensor Processing Units (TPUs), specifically the TPUv7, are emerging as a formidable alternative to Nvidia’s GPUs, which have dominated AI advancements for over a decade. This shift is exemplified by the training of cutting-edge models like Google’s Gemini 3 and Anthropic’s Claude 4.5 Opus on TPUs rather than Nvidia hardware.

A new framework called FedMining has been proposed to enhance underground mine safety through secure and privacy-preserving federated learning (FL). This approach allows for decentralized model training using sensor networks to monitor critical parameters, addressing privacy concerns associated with transmitting raw data to centralized servers.

A new study introduces a skewness-guided pruning method for multimodal Swin Transformers, aimed at enhancing federated skin lesion classification on edge devices. This method selectively prunes specific layers based on the statistical skewness of their output distributions, addressing the challenges of deploying large, computationally intensive models in medical imaging.

A recent study has introduced a new optimization approach in Federated Learning (FL) that minimizes Layerwise Activation Norm to enhance the generalization of models trained in a federated setup. This method addresses the issue of the global model converging to a 'sharp minimum', which can negatively impact its performance across diverse datasets. By imposing a flatness constraint on the Hessian derived from training loss, the study aims to improve model robustness and adaptability.