26 papers · ranked by Valyu relevance
Gang Liu, Yihan Zhu, Jie Chen, Meng Jiang
Large language models hold promise as scientific assistants, yet existing agents either rely solely on algorithm evolution or on deep research in isolation, both of which face critical limitations. Pure algorithm evolution, as in AlphaEvolve, depends only on the internal knowledge of LLMs and quickly plateaus in…
Leyi Zhao, Weijie Huang, Yitong Guo, Jiang Bian + 2 more
Optimizing scientific computing algorithms for modern GPUs remains a labor-intensive and iterative process, requiring repeated cycles of code modification, benchmarking, and tuning across complex hardware and software stacks. While recent work has explored large language model (LLM)–assisted evolutionary algorithms for…
Zhimeng Zhou, Yang Nan, Minjie Mou, Yuntao Qian + 16 more
Artificial intelligence (AI) is increasingly permeating the drug development pipeline. Numerous algorithms for accelerating this multi-stage and multi-task process have been constructed, which depends heavily on expert design and labor-intensive task-specific optimization. Given that AI-driven acceleration of drug…
Kaichen Ouyang, Yu, Mingyang, Ke + 11 more
This paper introduces a novel framework linking evolutionary computation to statistical physics by formulating optimization as a statistical phase transition. We propose Wasserstein Evolution (WE), an algorithm based on the Wasserstein gradient flow of a free energy functional, translating the physical competition…
Rongjie Liao, Junhao Qiu, Xin Chen, Xiaoping Li
Customized static operator design has enabled widespread application of Evolutionary Algorithms (EAs), but their search performance is transient during iterations and prone to degradation. Dynamic operators aim to address this but typically rely on predefined designs and localized parameter control during the search…
Minghao Yan, Bo Peng, Benjamin Coleman, Ziqi Chen + 10 more
Large language models have become drivers of evolutionary search, but most systems rely on a fixed, prompt-elicited policy to sample next candidates. This limits adaptation in practical engineering and research tasks, where evaluations are expensive, and progress depends on learning task-specific search dynamics. We…
Jiayi Zhang, Yongfeng Gu, Jianhao Ruan, Maojia Song + 9 more
Agentic evolution has emerged as a powerful paradigm for improving programs, workflows, and scientific solutions by iteratively generating candidates, evaluating them, and using feedback to guide future search. However, existing methods are typically instantiated either as fixed hand-designed procedures that are…
Irving Luna-Ortiz, Alejandro Rodríguez-Molina, Miguel Gabriel Villarreal-Cervantes, Mario Aldape-Pérez + 3 more
This paper introduces a variant of differential evolution with micro-populations, called $μ$-DE-ERM, which incorporates a periodic elitist replacement mechanism with the aim of preserving diversity without the need to measure it explicitly. The proposed algorithm is designed for scenarios with reduced evaluation…
Innocent Sibanda, Geoff Nitschke
The goal of bioengineering in synthetic biology is to redesign, reprogram, and rewire biological systems for specific applications using standardized parts such as promoters and ribosomes. For example, bioengineered micro-organisms capable of cleaning up environmental pollution or producing antibodies de novo to defend…
Inès Benito, Johannes F. Lutzeyer, Benjamin Doerr
Baldwinian and Lamarckian evolution have existed for a long time in evolutionary algorithms (EAs) without ever dominating the academic literature or practical applications. In this work, we use modern empirical and theoretical methods to revisit Lamarckian and Baldwinian evolution and rigorously compare them with the…
Steven A. Frank, Antonio M. Scarfone
Diverse learning algorithms, optimization methods, and natural selection share a common mathematical structure despite their apparent differences. Here, I show that a simple notational partitioning of change by the Price equation reveals a universal force-metric-bias (FMB) law: $Δθ=(Mf+b+ξ)$. The force $f$ drives…
Maxim Sakharov, Heming Jia
Memetic algorithms achieve strong optimization performance by combining population-based global search with local refinement operators, yet their effectiveness critically depends on the design and management of memes. Local search strategies are typically handcrafted, problem-specific, and fixed prior to execution.…
Yang Cao, Bingchuan Wu, Miao Wen, Yang Lou
Differential evolution has become one of the mainstream solvers for complex optimization problems due to its concise structure and strong global search ability. However, the performance of the DE algorithm is highly sensitive to its mutation and crossover strategies and related control parameters. Traditional adaptive…
Authors not listed
Finding the most stable adsorption geometry of a flexible molecule on a catalytic surface remains a key challenge due to the high dimensionality and ruggedness of the potential energy surface. We present a Gradient-Enhanced Genetic Algorithm (GE-GA) for the global optimization of adsorbate–surface configurations…
Authors not listed
Designing molecules with specific target properties remains a fundamental challenge in computational chemistry. While existing approaches show promise, most rely on simplified representations like SMILES strings or 2D graphs that lack essential three-dimensional geometric information. We present EvoDiffMol, a…
Xiaoyu Cao, Yueshun He, Linlin He, Wei Lv + 2 more
3D UAV path planning intends to produce safe, feasible, and optimized flight trajectories in complex 3D environments containing multiple obstacles and coupled constraints. Nevertheless, many existing optimization approaches encounter difficulties when dealing with high-dimensional constrained search spaces, since…
Tshegofatso Botlhoko, Tlhalitshi Volition Montshiwa
Background One of the disadvantages of the multilayer perception (MLP), which is a machine learning (ML) algorithm used in various fields, includes the uncontrollable growth of the number of total parameters, which may make MLP redundant in such high dimensions, and the uncontrollable growing stack of layers that…
Andrea Polo-Rodríguez, David R. Penas, Julio R. Banga
Parameter estimation is a central challenge in systems biology, particularly for large dynamic models described by nonlinear ordinary differential equations (ODEs). These global optimization problems exhibit landscapes which are topologically heterogeneous, often exhibiting a pathological mixture of stiff, smooth…
Authors not listed
Machine olfaction—the artificial replication of the sense of smell—faces significant challenges due to the absence of large, standardized training datasets. Unlike vision, language, and audio models, which benefit from extensive corpora such as ImageNet, GLUE, and AudioSet, olfaction lacks scaled equivalents and…
Iain G. Johnston, Ramon Diaz-Uriarte, James D. Boyko
Many scientific questions involve the coevolution of coupled, binary features over time – from phenotypes in evolutionary biology to mutations in cancer development. Evolutionary accumulation models (EvAMs) often neglect reversibility in these systems, uncertainty in observations, and/or phylogenetic connections…
Authors not listed
Mesoporous adsorbent materials offer a large volumetric capacity; however, cyclic adsorption/desorption processes in these systems often suffer from hysteresis and may require a significant pressure swing to access this capacity. To mitigate hysteresis, a proposed strategy is to include nucleation sites on the walls of…
Nhan Ly-Trong, Samuel Martin, Nick Goldman, Nicola De Maio + 1 more
Phylogenetic analysis is essential to genomic epidemiology, for example in tracing the origin and evolution of SARS-CoV-2 variants during the COVID-19 pandemic. We previously introduced CMAPLE, a single-threaded implementation of the MAPLE algorithm designed for large-scale epidemiological genomic datasets. CMAPLE can…
Changin Oh, Kathleen P. Wilkie
We present the Toroidal Search Algorithm (TSA), a novel population-based metaheuristic optimization method inspired by the topology of a torus. Conventional metaheuristics frequently suffer from boundary stagnation, a phenomenon that severely degrades performance in bounded and high-dimensional search spaces. TSA…
Annabel Large, Ian Holmes
Most statistical phylogenetics analyses use relatively simple continuous-time finite-state Markov models of point substitution to describe molecular evolution, often keeping sequence length fixed, ignoring insertions and deletions (indels) entirely, and making little (if any) allowance for variations in selection…
Authors not listed
Incorporating prior domain knowledge into Bayesian optimization (BO) remains difficult for statistical methods, which also typically suffer from limited interpretability. Large language models (LLMs) offer complementary strengths in reasoning and knowledge integration, but it remains unclear when and how they improve…
Authors not listed
Metastable states and the conformational transitions in between them are key to understanding dynamical behaviour and function of large-scale molecular systems. By combining basic dimensionality reduction techniques with a state-of-the art approximation of the Koopman operator associated to molecular dynamics…