Browse Articles
Discover research articles across all indexed journals
First ‘half-Möbius’ carbon chain wows chemists
A novel automatic modulation recognition algorithm for OFDM signals based on FAFT
25-Hydroxyvitamin D3 promotes slow-twitch fiber type transition in skeletal muscle
The first ‘AI societies’ are taking shape: how human-like are they?
Bioremediation of lunar regolith simulant through mycorrhizal fungi and plant symbioses enables chickpea to seed
Abstract Food sustainability is a significant challenge for long-term space travel. Plants can provide fresh nutrition, reducing reliance on packaged foods. Using Lunar regolith simulant (LRS), we tested a methodology to create a productive growth medium for horticultural crops on the Moon. We leveraged chickpea ( Cicer arietinum ), Arbuscular Mycorrhizal Fungi (AMF), and Vermicompost (VC) to enhance plant stress tolerance, sequester contaminants, and improve substrate structure. Chickpeas were cultivated in LRS/VC mixtures, with or without AMF, under climate-controlled conditions. Plants seeded successfully in mixtures containing up to 75% LRS when inoculated with AMF. While the number of seeds declined with increasing LRS concentration, seed size remained stable. Higher LRS concentrations induced stress; however, plants grown in 100% LRS inoculated with AMF demonstrated an average extension of two weeks in survival compared to non-inoculated plants. AMF colonized roots across all mixtures, including 100% LRS, demonstrating the ability to establish symbioses under extreme conditions. We also observed improvement in the structural properties of LRS by forming aggregates capable of withstanding extreme conditions, potentially mitigating particle-related hazards. These results provide a baseline for chickpea establishment and yield in amended LRS while demonstrating biological improvements in regolith properties.
Bi-level graph attention paradigm with differential strategy integration for heterogeneous multi-agent reinforcement learning
Abstract Collaboration among heterogeneous agents is crucial for addressing complex real-world tasks that require leveraging diverse capabilities. In such systems, increasing agent numbers amplify the challenges of communication and coordinated decision-making, in addition to the inherent heterogeneity of the agents. To address these issues, we propose the Bi-level Graph Attention Paradigm (Bi-GAP) with differential strategy integration, a novel policy-based group learning framework designed for heterogeneous Multi-Agent Systems (MAS) in both discrete and continuous domains. Bi-GAP employs a bi-level graph attention architecture to model intricate interaction patterns among isomorphic agents within groups and across heterogeneous groups. This hierarchical representation enables flexible and selective communication, reduces unnecessary message exchange, and improves the robustness of the MAS under interference. Furthermore, the framework integrates multi-perspective strategies, allowing each member-agent to incorporate global guidance from its designated guide-agent while still performing fine-grained local reasoning. This mechanism balances macro-level coordination with micro-level adaptability. We evaluate Bi-GAP on heterogeneous StarCraft II micromanagement tasks and Multi-Agent Particle Environment Predator–Prey scenarios. The results show that Bi-GAP consistently outperforms recent state-of-the-art MARL baselines across both discrete and continuous settings.