631 papers
This collection explores the fascinating middle section of the periodic table, where elements like cesium and neodymium play critical roles in everything from high-precision atomic clocks to powerful magnets. These substances are not just laboratory curiosities; they drive modern technology, enabling advancements in telecommunications, renewable energy storage, and quantum computing research.
At Gist.Science, we monitor arXiv daily to ensure you never miss a breakthrough in this dynamic area. We process every new preprint in this category, transforming complex research into both plain-language overviews and detailed technical summaries so you can grasp the science without getting lost in the jargon.
Below are the latest papers in this field, offering fresh insights into how these versatile elements are reshaping our technological future.
By exhaustively mapping a morphology-fitness landscape of over 1.3 million soft robots, this study reveals that while evolutionary brain-body co-optimization can yield unique performance gains through goal-switching, it consistently fails to select for morphological potential because it frequently undervalues and eliminates promising newly mutated bodies.
This paper presents a proof-of-concept demonstration of "morphological cognition," showing that a robot constructed from simulated voxels with fixed behaviors can classify MNIST digits (specifically distinguishing 0s from 1s) and move accordingly without employing any neural circuitry.
This paper demonstrates that Evolution Strategies (ES) outperform Reinforcement Learning (RL) in post-training Large Language Models for discovery domains by maintaining broader solution coverage and higher pass@k scores, thereby avoiding the output distribution collapse that limits RL's effectiveness in generating diverse candidate solutions.
This paper proposes a Deep Asymmetric Ternary Spiking Q-Network (DATSQN) that utilizes a novel ternary spiking neuron model to mitigate gradient estimation bias, thereby overcoming the performance degradation of existing ternary models and outperforming binary baselines in deep Q-learning tasks across seven Atari games.
This paper proposes Adaptive Hybrid Particle Swarm Optimization (AHPSO), a method that automatically modulates gradient injection based on swarm diversity to enhance performance on smooth local basins, demonstrating that while it does not universally outperform standard PSO, it achieves superior rankings against top algorithms like CMA-ES under iteration-matched comparisons.
The paper introduces Lapis, a multiplication-free spiking attention mechanism that leverages first-spike latency differences and a Laplacian kernel to achieve near-state-of-the-art accuracy on CIFAR-10 and ImageNet-1K while significantly reducing arithmetic energy consumption.
This paper demonstrates that homogeneous pivoting cube modular robots can achieve effective global reconfiguration in two dimensions using only local sensing and reinforcement learning, where incorporating grid symmetries reduces model size and multi-step information passing enables near-optimal performance despite limited neighbor interactions.
This paper proposes a Graph Neural Network-guided Genetic Algorithm (GNN-GA) to optimize Physical Internet supply chain planning under cost uncertainty by leveraging learned hub-specific factory-selection probabilities for initialization and uncertainty-aware mutation, demonstrating superior performance over standard genetic algorithms and simulated annealing in solving complex three-echelon network assignment problems.
EvoMem is a persistent memory architecture for LLM-driven evolutionary code search that captures and reuses successful mutation strategies as structured, task-aware advice to reduce redundant exploration and improve optimization across diverse tasks.
This paper demonstrates that deep binary and sparse ternary ReLU networks can effectively approximate -Hölder functions and achieve the minimax prediction rate for -smooth regression, up to a logarithmic factor.