The European Commission has announced a funding opportunity titled "Enhanced Learning Strategies for General Purpose AI: Advancing GenAI4EU," aiming to advance General Purpose AI (GPAI) models beyond current capabilities. While existing large-scale AI models excel in tasks like natural language processing and image generation, they face challenges in adaptability, complex reasoning, and self-awareness of limitations. This call seeks to develop GPAI models that integrate self-supervised learning with complementary strategies such as hybrid learning, active learning, reinforcement learning, relational learning, continual learning, evolutionary learning, and physics-based learning. The objective is to create robust, adaptive AI systems capable of formal reasoning, mathematical problem-solving, confidence estimation, long-term planning, and seamless adaptation to dynamic environments.

Proposals should focus on at least one of the following research areas:

• Hybrid Learning Architectures for Advanced Reasoning: Combining self-supervised learning with symbolic reasoning and neuro-symbolic methods to enhance reasoning and problem-solving abilities.
• Continual and Evolutionary Learning for Dynamic Environments: Enabling GPAI models to adapt and learn from changing conditions, retaining essential knowledge for real-world operations.
• Reinforcement Learning Integration: Merging self-supervised learning with reinforcement learning to address challenges like non-stationary data and computational costs.
• Explainable AI and Trustworthy Decision-Making: Incorporating explainable AI methodologies, including causal inference and counterfactual reasoning, to enhance transparency and accountability.
• Other Novel Paradigms: Exploring combinations of self-supervised learning with paradigms like active learning, relational learning, and embodied learning to equip GPAI models with advanced capabilities.

Projects should balance theoretical advancements with practical applications, aligning with European values and principles, including the AI Act. The research aims to impact sectors such as advanced robotics, personalized healthcare, mobility, manufacturing, sustainable energy solutions, and scientific research. Interdisciplinary collaboration is encouraged to ensure comprehensive assessments and responsive tool development. Proposals must adhere to Horizon Europe's Open Science practices, providing open access to research outputs unless constrained by legitimate reasons. Mechanisms for assessing progress, including qualitative and quantitative KPIs, benchmarking, and progress monitoring, should be incorporated. Results should be shared with the European R&D community through platforms like the AI-on-demand platform to bolster the European AI, Data, and Robotics ecosystem. Collaboration with existing projects and synergies with other European, national, or regional initiatives is also expected.

Opening: 10 Jun 2025

Deadline(s): 02 Oct 2025