Evolutionary Optimization

Ask about this topic

Overview

Evolutionary optimization studies how populations of candidate solutions can improve through selection, mutation, recombination, and other forms of variation, especially when gradients are unavailable or objectives are expensive. Common approaches include genetic algorithms, differential evolution, CMA-ES and evolution strategies, as well as multi-objective methods that preserve trade-offs and diversity. Researchers apply them to engineering design, robotics, scheduling, scientific discovery, and increasingly to evolving code, prompts, and language-model behavior; newer methods add learned surrogates, adaptive resource allocation, and large-language-model-generated operators. A central challenge is balancing exploration, solution quality, and evaluation cost, while ensuring improvements generalize beyond the benchmarks and search budgets used to find them.

Latest papers

CardsList

295 papers33 papers in the last four weeks, up 50% on the four weeks before. 0.2% of all new papers.

Feb 26, 2026

EvoX: Meta-Evolution for Automated Discovery

Recent work such as AlphaEvolve has shown that combining LLM-driven optimization with evolutionary search can effectively improve programs, prompts, and algorithms across domains. In this paradigm, previously evaluated solutions are reused to guide the model toward new candidate solutions. Crucially, the effectiveness of this evolution process depends on the search strategy: how prior solutions are selected and varied to generate new candidates. However, most existing methods rely on fixed search strategies with predefined knobs (e.g., explore-exploit ratios) that remain static throughout execution. While effective in some settings, these approaches often fail to adapt across tasks, or even within the same task as the search space changes over time. We introduce EvoX, an adaptive evolution method that optimizes its own evolution process. EvoX jointly evolves candidate solutions and the search strategies used to generate them, continuously updating how prior solutions are selected and varied based on progress. This enables the system to dynamically shift between different search strategies during the optimization process. Across nearly 200 real-world optimization tasks, EvoX outperforms existing AI-driven evolutionary methods including AlphaEvolve, OpenEvolve, GEPA, and ShinkaEvolve on the majority of tasks.

Shu Liu, Shubham Agarwal, Monishwaran Maheswaran +14

UC Berkeley · Stanford University · Bespoke Labs

Evolutionary OptimizationLLMs for Evolutionary OptimizationLarge Language Model-Guided OptimizationAutomated Algorithm Discovery
AskPDF
Feb 25, 2026

Duel-Evolve: Reward-Free Test-Time Scaling via LLM Self-Preferences

Many applications seek to optimize LLM outputs at test time by iteratively proposing, scoring, and refining candidates over a discrete output space. Existing methods use a calibrated scalar evaluator for the target objective to guide search, but for many tasks such scores are unavailable, too sparse, or unreliable. Pairwise comparisons, by contrast, are often easier to elicit, still provide useful signal on improvement directions, and can be obtained from the LLM itself without external supervision. Building on this observation, we introduce Duel-Evolve, an evolutionary optimization algorithm that replaces external scalar rewards with pairwise preferences elicited from the same LLM used to generate candidates. Duel-Evolve aggregates these noisy candidate comparisons via a Bayesian Bradley-Terry model, yielding uncertainty-aware estimates of candidate quality. These quality estimates guide allocation of the comparison budget toward plausible optima using Double Thompson Sampling, as well as selection of high-quality parents to generate improved candidates. We evaluate Duel-Evolve on MathBench, where it achieves 20 percentage points higher accuracy over existing methods and baselines, and on LiveCodeBench, where it improves over comparable iterative methods by over 12 percentage points. Notably, the method requires no reward model, no ground-truth labels during search, and no hand-crafted scoring function. Results show that pairwise self-preferences provide strong optimization signal for test-time improvement over large, discrete output spaces.

Sweta Karlekar, Carolina Zheng, Magnus Saebo +5

Columbia University · Los Alamos National Laboratory

Large Language Model-Guided OptimizationTest-Time ScalingLanguage Model Self-ImprovementEvolutionary OptimizationLanguage Model Self-Assessment
AskPDF
Feb 25, 2026

Code World Models for Parameter Control in Evolutionary Algorithms

Can an LLM learn how an optimizer behaves -- and use that knowledge to control it? We extend Code World Models (CWMs), LLM-synthesized Python programs that predict environment dynamics, from deterministic games to stochastic combinatorial optimization. Given suboptimal trajectories of (1+1)(1{+}1)-RLSk\text{RLS}_k, the LLM synthesizes a simulator of the optimizer's dynamics; greedy planning over this simulator then selects the mutation strength kk at each step. On \lo{} and \onemax{}, CWM-greedy performs within 6% of the theoretically optimal policy -- without ever seeing optimal-policy trajectories. On \jump{k_k}, where a deceptive valley causes all adaptive baselines to fail (0% success rate), CWM-greedy achieves 100% success rate -- without any collection policy using oracle knowledge of the gap parameter. On the NK-Landscape, where no closed-form model exists, CWM-greedy outperforms all baselines across fifteen independently generated instances (36.9436.94 vs.\ 36.3236.32; p<0.001p<0.001) when the prompt includes empirical transition statistics. The CWM also outperforms DQN in sample efficiency (200 offline trajectories vs.\ 500 online episodes), success rate (100% vs.\ 58%), and generalization (k=3k{=}3: 78% vs.\ 0%). Robustness experiments confirm stable synthesis across 5 independent runs.

Camilo Chacón Sartori, Guillem Rodríguez Corominas

ICN2 – Catalan Institute of Nanoscience and Nanotechnology, Barcelona, Spain · Artificial Intelligence Research Institute (IIIA-CSIC), Barcelona, Spain · Universitat Politècnica de Catalunya (UPC), Barcelona, Spain

Code World ModelsLarge Language Model-Guided OptimizationLLMs for Evolutionary OptimizationWorld ModelsEvolutionary OptimizationCombinatorial Optimization
AskPDF
Feb 23, 2026

AdaEvolve: Adaptive LLM Driven Zeroth-Order Optimization

The paradigm of automated program generation is shifting from one-shot generation to inference-time search, where Large Language Models (LLMs) function as semantic mutation operators within evolutionary loops. While effective, these systems are currently governed by static schedules that fail to account for the non-stationary dynamics of the search process. This rigidity results in substantial computational waste, as resources are indiscriminately allocated to stagnating populations while promising frontiers remain under-exploited. We introduce AdaEvolve, a framework that reformulates LLM-driven evolution as a hierarchical adaptive optimization problem. AdaEvolve uses an "accumulated improvement signal" to unify decisions across three levels: Local Adaptation, which dynamically modulates the exploration intensity within a population of solution candidates; Global Adaptation, which routes the global resource budget via bandit-based scheduling across different solution candidate populations; and Meta-Guidance which generates novel solution tactics based on the previously generated solutions and their corresponding improvements when the progress stalls. We demonstrate that AdaEvolve consistently outperforms the open-sourced baselines across 185 different open-ended optimization problems including combinatorial, systems optimization and algorithm design problems.

Mert Cemri, Shubham Agrawal, Akshat Gupta +9

University of California, Berkeley · Bespoke Labs

Large Language Model-Guided OptimizationLLMs for Evolutionary OptimizationEvolutionary OptimizationLLM-Based Program SynthesisZeroth-Order Optimization
AskPDF
Feb 22, 2026

All Constant Mutation Rates for the (1+1)(1+1) Evolutionary Algorithm

For every mutation rate p∈(0,1)p \in (0, 1), and for all ε>0\varepsilon > 0, there is a fitness function f:{0,1}n→Rf : \{0,1\}^n \to \mathbb{R} with a unique maximum for which the optimal mutation rate for the (1+1)(1+1) evolutionary algorithm on ff is in (p−ε,p+ε)(p-\varepsilon, p+\varepsilon). In other words, the set of optimal mutation rates for the (1+1)(1+1) EA is dense in the interval [0,1][0, 1]. To show that, this paper introduces DistantSteppingStones, a fitness function which consists of large plateaus separated by large fitness valleys.

Andrew James Kelley

Fitness Landscape AnalysisEvolutionary OptimizationRuntime Analysis of Evolutionary Algorithms
AskPDF
Feb 18, 2026

RankEvolve: Automating the Discovery of Retrieval Algorithms via LLM-Driven Evolution

Retrieval algorithms like BM25 and query likelihood with Dirichlet smoothing remain strong and efficient first-stage rankers, yet improvements have mostly relied on parameter tuning and human intuition. We investigate whether a large language model, guided by an evaluator and evolutionary search, can automatically discover improved lexical retrieval algorithms. We introduce RankEvolve, a program evolution setup based on AlphaEvolve, in which candidate ranking algorithms are represented as executable code and iteratively mutated, recombined, and selected based on retrieval performance across 12 IR datasets from BEIR and BRIGHT. RankEvolve starts from two seed programs: BM25 and query likelihood with Dirichlet smoothing. The evolved algorithms are novel, effective, and show promising transfer to the full BEIR and BRIGHT benchmarks as well as TREC DL 19 and 20. Our results suggest that evaluator-guided LLM program evolution is a practical path towards automatic discovery of novel ranking algorithms.

Jinming Nian, Fangchen Li, Dae Hoon Park +1

Santa Clara University, Santa Clara, CA, USA · Independent Researcher, Bothell, WA, USA · Walmart Global Tech, Sunnyvale, CA, USA

Information RetrievalAutomated Algorithm DiscoveryLexical RetrievalLarge Language Model-Guided OptimizationEvolutionary Optimization
AskPDF
Feb 17, 2026

MadEvolve: Evolutionary Optimization of Cosmological Algorithms with Large Language Models

We develop a general framework to discover scientific algorithms and apply it to three problems in computational cosmology. Our code, MadEvolve, is similar to Google's AlphaEvolve, but places a stronger emphasis on free parameters and their optimization. Our code starts with a baseline human algorithm implementation, and then optimizes its performance metrics by making iterative changes to its code. As a further convenient feature, MadEvolve automatically generates a report that compares the input algorithm with the evolved algorithm, describes the algorithmic innovations and lists the free parameters and their function. Our code supports both auto-differentiable, gradient-based parameter optimization and gradient-free optimization methods. We apply MadEvolve to the reconstruction of cosmological initial conditions, 21cm foreground contamination reconstruction and effective baryonic physics in N-body simulations. In all cases, we find substantial improvements over the base algorithm. We make MadEvolve and our three tasks publicly available at madevolve.org.

Tianyi Li, Shihui Zang, Moritz Münchmeyer

Department of Physics, University of Wisconsin-Madison

Automated Algorithm DiscoveryScientific Code GenerationEvolutionary OptimizationLarge Language Model-Guided Scientific Discovery
AskPDF
Feb 16, 2026

DexEvolve: Evolutionary Optimization for Robust and Diverse Dexterous Grasp Synthesis

Dexterous grasping is fundamental to robotics, yet data-driven grasp prediction heavily relies on large, diverse datasets that are costly to generate and typically limited to a narrow set of gripper morphologies. Analytical grasp synthesis can be used to scale data collection, but necessary simplifying assumptions often yield physically infeasible grasps that need to be filtered in high-fidelity simulators, significantly reducing the total number of grasps and their diversity. We propose a scalable generate-and-refine pipeline for synthesizing large-scale, diverse, and physically feasible grasps. Instead of using high-fidelity simulators solely for verification and filtering, we leverage them as an optimization stage that continuously improves grasp quality without discarding precomputed candidates. More specifically, we initialize an evolutionary search with a seed set of analytically generated, potentially suboptimal grasps. We then refine these proposals directly in a high-fidelity simulator (Isaac Sim) using an asynchronous, gradient-free evolutionary algorithm, improving stability while maintaining diversity. In addition, this refinement stage can be guided toward human preferences and/or domain-specific quality metrics without requiring a differentiable objective. We further distill the refined grasp distribution into a diffusion model for robust real-world deployment, and highlight the role of diversity for both effective training and during deployment. Experiments on a newly introduced Handles dataset and a DexGraspNet subset demonstrate that our approach achieves over 120 distinct stable grasps per object (a 1.7-6x improvement over unrefined analytical methods) while outperforming diffusion-based alternatives by 46-60% in unique grasp coverage.

René Zurbrügg, Andrei Cramariuc, Marco Hutter

ETH Zürich

Robotic GraspingDexterous ManipulationSynthetic Data Generation for RoboticsEvolutionary Optimization
AskPDF
Feb 16, 2026

Evolutionary System Prompt Learning for Reinforcement Learning in LLMs

Building agentic systems that can autonomously self-improve from experience is a longstanding goal of AI. Large language models (LLMs) today primarily self-improve via two mechanisms: self-reflection for context updates, and reinforcement learning (RL) for weight updates. In this work, we propose Evolutionary System Prompt Learning (E-SPL), a method for jointly improving model contexts and model weights. In each RL iteration, E-SPL samples trajectories under multiple system prompts in parallel, then jointly applies RL updates to LLM weights and evolutionary updates to system prompts. System prompts evolve via mutation and crossover, two genetic operators driven by LLM self-reflection; selection is based on relative performance ratings updated across RL iterations. E-SPL encourages a natural division between declarative knowledge encoded in prompts and procedural knowledge encoded in weights, resulting in improved performance across reasoning and agentic tasks. For instance, in an easy-to-hard (AIME →\rightarrow BeyondAIME) generalization setting, E-SPL improves RL success rate from 38.8% →\rightarrow 45.1% while also outperforming reflective prompt evolution (40.0%). Overall, our results demonstrate that RL and system prompt evolution are deeply synergistic, and combining the two yields consistent gains in sample efficiency and generalization. Code: https://github.com/LunjunZhang/E-SPL

Lunjun Zhang, Ryan Chen, Bradly C. Stadie

Department of Computer Science, University of Toronto · Department of Statistics and Data Science, Northwestern University · Bridgewater AIA Labs

Language Model Self-ImprovementRL for Language ModelsRL for Language Model ReasoningEvolutionary Optimization
AskPDF
Feb 13, 2026

NeuroWeaver: An Autonomous Evolutionary Agent for Exploring the Programmatic Space of EEG Analysis Pipelines

Although foundation models have achieved remarkable success in general domains, applying them to electroencephalography (EEG) analysis is constrained by substantial data requirements and large parameter counts, which incur prohibitive computational costs and impede deployment in resource-constrained clinical environments. General-purpose automated machine learning frameworks are likewise ill-suited to this domain, since exploration within an unbounded programmatic space fails to incorporate essential neurophysiological priors and frequently yields neuroscientifically implausible solutions. We therefore propose NeuroWeaver, a unified autonomous evolutionary agent that generalizes across diverse EEG datasets and tasks by reformulating pipeline engineering as a discrete constrained optimization problem solved through large language model (LLM)-driven generation of executable code. A Domain-Informed Subspace Initialization confines the search to a neuroscientifically plausible manifold, while a Multi-Objective Evolutionary Optimization dynamically balances performance, novelty, and efficiency via self-reflective refinement. Across five heterogeneous benchmarks, NeuroWeaver synthesizes lightweight pipelines that outperform state-of-the-art task-specific methods on nearly all metrics and attain accuracy comparable to large-scale foundation models, even surpassing them on the HMC and Workload benchmarks with only 0.180.18M and 0.0110.011M parameters, respectively.

Guoan Wang, Shihao Yang, Feng Liu

Department of Industrial and Systems Engineering, Rutgers, The State University of New Jersey, 96 Frelinghuysen Road, Piscataway, NJ 08854, USA

Evolutionary OptimizationEEG DecodingLarge Language Model-Guided OptimizationLLM-Based Program SynthesisAutomated Machine Learning
AskPDF
Feb 10, 2026

ImprovEvolve: Basin-Hopping Meets LLM-Guided Evolutionary Search

LLM-guided evolutionary computation, most notably AlphaEvolve, has been remarkably successful in discovering novel mathematical constructions by solving challenging optimization problems. The standard approach is to evolve a monolithic program that directly outputs a candidate solution. We present ImprovEvolve, an algorithmic alternative that drastically reduces cognitive load on the LLM. Instead of prompting the model for an end-to-end optimizer, we evolve a program with three specialized operators of initialization, local improvement, and perturbation. We then approach the optimum by iteratively applying local improvements and intensity-scheduled perturbations, effectively driving a basin-hopping search with LLM-evolved subroutines. For hexagon in hexagon packing, ImprovEvolve discovers new state-of-the-art packings of 11, 12, 15, and 16 hexagons, and additionally for 14, 17, and 23 hexagons after minimal expert tuning of the generated code. For the second autocorrelation inequality, the evolved and human-scaled program pushes the lower bound from 0.96102 to 0.96258. For spherical codes, the ImprovEvolve program lowers the best-known maximum cosine for the majority of 90 randomly chosen diverse state-of-the-art spherical codes, achieving relative improvements of up to 2.4%.

Alexey Kravatskiy, Valentin Khrulkov, Ivan Oseledets

1MIRIAI, Russia · 2FusionBrain Lab, Russia · 3AXXX, Russia +1

Large Language Model-Guided Scientific DiscoveryEvolutionary OptimizationMetaheuristic OptimizationLarge Language Model-Guided OptimizationCombinatorial OptimizationGlobal Optimization+2
AskPDF
Feb 1, 2026

Organisms Write Their Own Evolution: Direct Epigenomic Inheritance Accelerates Learning on Compositional Tasks

What if organisms could write their own evolution? For 150 years, evolution has been understood as one-way: genes provide the rules, organisms respond to selection. But if organisms could learn during their lifetime, write that learned knowledge directly into their epigenome -- the heritable regulatory layer controlling gene expression -- and pass it to offspring, then knowledge would accumulate across generations instead of being rediscovered by each one. Such direct epigenomic inheritance would constitute genuine organismal agency: writing new information into heredity rather than passively receiving it. To test whether organisms actually benefit from this capability, we implement Phenopoiesis, a computational model in which organisms maintain two independent heritable layers: a genetic layer (inherited with mutation) and an epigenetic layer (inherited unchanged). We compare Phenopoiesis against Selection Only (no learning) and the Baldwin Effect (learning influences selection but is not inherited) across six learning tasks of increasing complexity: three simple (learn one pattern) and three compositional (learn two patterns simultaneously). On simple tasks all three mechanisms perform equivalently (p = 0.559): complexity is required for inheritance to matter. On learnable compositional tasks, Phenopoiesis outperforms the alternatives by 3-6x in fitness (F = 177-358, p < 0.001, Cohen's d = 3.44-4.89), with perfect reproducibility (s.d. = 0.0 on T+Plus across all 30 runs), indicating deterministic knowledge accumulation: offspring inherit their parents' learned solutions and add new learning on top. On a geometrically incompatible task all mechanisms fail equally, revealing an inherent limit. These results give the first direct computational test of Denis Noble's hypothesis that organisms possess read-write access to their own heredity, and show precisely when that access pays off.

Nam H. Le

Department of Computer Science University of Vermont

Evolutionary Optimization
AskPDF
Jan 27, 2026

Online Regime-aware Calibration for Black-box Social Simulators via Posterior-assisted Evolutionary Dynamic Optimization

Evolutionary dynamic optimization (EDO) commonly assumes that environmental changes can be detected from fitness variations and handled through random re-initialization, historical solutions, or learned transition patterns. Online calibration of black-box simulators introduces a different setting, where the dynamic objective is induced by sequential observations and a changing calibration window, rather than being controlled by explicit variables. Fitness variations therefore cannot be directly attributed to regime changes, while the unknown relationship between successive regimes limits conventional adaptation. We formulate this setting as an observation-driven dynamic optimization problem and propose PosEDO, which augments fitness-based EDO with an observation-conditioned parameter-space signal. PosEDO learns this signal online as a posterior distribution over simulator parameters from parameter-trajectory pairs generated during evolutionary evaluation, using posterior shifts for change detection and posterior samples for population adaptation. The new evaluation records are further utilized for online posterior updating without additional simulator calls. Experiments on nonstationary economic and financial simulators show that PosEDO improves calibration accuracy, optimization performance, and change-detection quality over representative EDO baselines.

Peng Yang, Zhenhua Yang, Boquan Jiang +3

Guangdong Provincial Key Laboratory of Brain-Inspired Intelligent Computation, Southern University of Science and Technology, Shenzhen 518055, China · Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen 518055, China · Department of Statistics and Data Science, Southern University of Science and Technology, Shenzhen 518055, China +2

Evolutionary OptimizationBlack-Box OptimizationModel Calibration
AskPDF
Jan 19, 2026

An Evolutionary Framework for Automatic Optimization Benchmark Generation via Large Language Models

Optimization benchmarks play a fundamental role in assessing algorithm performance; however, existing artificial benchmarks often fail to capture the diversity and irregularity of real-world problem structures, while benchmarks derived from real-world problems are costly and difficult to construct. To address these challenges, we propose an evolutionary automatic benchmark generation framework that leverages a large language model (LLM) as a generative operator, termed the LLM-driven evolutionary benchmark generator (LLM-EBG). In this framework, the LLM serves as an evolutionary operator that generates and evolves benchmark problems within a flexible, expressive representation space. As a case study, we generate unconstrained single-objective continuous minimization problems represented as mathematical expressions designed to induce significant performance differences between a genetic algorithm (GA) and differential evolution (DE). Experimental results show that LLM-EBG successfully produces benchmark problems in which the designated target algorithm consistently outperforms the comparative algorithm in more than 80% of trials. Furthermore, exploratory landscape analysis reveals that benchmarks favoring GA are highly sensitive to variable scaling, demonstrating that the proposed framework can generate problems with distinct geometric characteristics that reflect the intrinsic search behaviors of different optimization algorithms.

Yuhiro Ono, Tomohiro Harada, Yukiya Miura

Graduate School of Systems Design, Tokyo Metropolitan University, 6-6, Asahigaoka, Hino, 1910065, Tokyo, Japan. · Graduate School of Science and Engineering, Saitama University, 2552026 Shimo-okubo, Sakura-ku, Saitama, 3388570, Saitama, Japan.

Evolutionary OptimizationBenchmark DesignSynthetic Benchmark Generation
AskPDF
Jan 15, 2026

PACEvolve: Enabling Progress-Aware Consistent Evolution

Self-evolving agents powered by Large Language Models (LLMs) have emerged as a promising direction across diverse domains, including code optimization and scientific discovery, yet their core failure modes remain underexplored. Through a comprehensive empirical study, we identify that the model's reasoning becomes anchored to the local context of current hypotheses, overemphasizing low-level details while neglecting the broader search landscape. As a result, such agents become prone to context pollution and mode collapse, repeatedly revisiting flawed hypotheses and converging on suboptimal solutions. To address this challenge, we propose Progress-Aware Consistent Evolution (PACEvolve), a systematic framework for governing agent memory and search dynamics. PACEvolve overcomes these limitations through three key techniques: (1) Hierarchical Context Management (HCM), which structures historical trajectories while dynamically pruning branches to preserve a high-signal memory state; (2) Momentum-Based Backtracking (MBB), which monitors optimization progress to escape local minima; and (3) a self-adaptive Collaborative Evolution policy (CE) that balances intra-trajectory refinement with inter-trajectory knowledge transfer. By decoupling high-level idea generation from low-level code evaluation, PACEvolve maintains a global view of search momentum and achieves state-of-the-art results across complex evolutionary benchmarks.

Minghao Yan, Bo Peng, Benjamin Coleman +13

University of Wisconsin–Madison · Google · Google DeepMind +1

Evolutionary OptimizationLarge Language Model-Guided OptimizationCode OptimizationSelf-Evolving AgentsMemory-Augmented Language Model Agents
AskPDF
Dec 30, 2025

Decoupling Constraints from Two Directions for Evolutionary Constrained Multi-objective Optimization

Real-world constrained multi-objective optimization problems (CMOPs) commonly involve multiple constraints, and understanding and exploiting their coupling relationships is crucial for efficient optimization. Recent constraint-decoupling methods handle individual constraints separately, but they generally search only in the evolutionary direction to approximate single-constraint Pareto fronts (SCPFs). In this study, we show that part or all of the constrained Pareto front (CPF) may be unrelated to any SCPF and instead be shaped by the boundaries of infeasible regions. We refer to such a portion as the independent CPF (ICPF) and introduce the reverse CPF (RCPF) to characterize its associated informative infeasible boundaries. Based on these observations, we propose a bidirectional constraint-decoupling coevolutionary algorithm named DCF2D. DCF2D dynamically identifies the constraints obstructing the main population and activates constraint-specific auxiliary populations. These populations adaptively search in the evolutionary direction for the corresponding SCPFs or in the reverse evolutionary direction for the corresponding RCPFs. Its three-stage framework integrates unconstrained global exploration, event-driven bidirectional coevolution, and final convergence refinement. Experiments on 87 benchmark instances from seven test suites and 28 real-world engineering CMOPs demonstrate that DCF2D achieves the best overall performance among nine algorithms. Code available at: https://github.com/RuiqingS/DCF2D.

Ruiqing Sun, Dawei Feng, Xing Zhou +6

College of Computer Science and Technology, National University of Defense Technology, Changsha 410000, P.R. China · College of Intelligence Science and Technology, National University of Defense Technology, Changsha, 410000, P.R. China · State Key Laboratory of Complex & Critical Software Environment, College of Information and Communication, National University of Defense Technology, Wuhan 430019, P.R. China +1

Evolutionary OptimizationMulti-Objective Evolutionary OptimizationConstrained OptimizationMulti-Objective Optimization
AskPDF
Nov 21, 2025

PepEVOLVE: Position-Aware Dynamic Peptide Optimization via Group-Relative Advantage

Macrocyclic peptides are an emerging modality that combines biologics-like affinity with small-molecule-like developability, yet their vast combinatorial space and multi-parameter objectives make lead optimization slow and challenging. Prior generative approaches for peptide lead optimization (e.g., PepINVENT) require chemists to pre-specify mutable positions for optimization, choices that are not always known a priori, and rely on static pretraining and optimization algorithms that limit the model's ability to generalize and effectively optimize peptide sequences. We introduce PepEVOLVE, a position-aware, dynamic framework that learns both where to edit and how to dynamically optimize peptides for multi-parameter improvement. PepEVOLVE (i) augments pretraining with dynamic masking and CHUCKLES shifting to improve generalization, (ii) uses a context-free multi-armed-bandit router that discovers high-reward residues, and (iii) couples a novel evolving optimization algorithm with group-relative advantage to stabilize reinforcement updates. During in silico evaluations, the router policy reliably learns and concentrates probability on chemically tractable sites that influence the desired peptide's properties. On a therapeutically motivated Rev-binding macrocycle benchmark, PepEVOLVE outperformed PepINVENT by reaching higher averaged permeability-weighted optimization scores of approximately 0.8 vs. 0.6, achieving best candidates with score of 0.97 (vs. 0.89), and converging in fewer steps under the task of optimizing permeability and lipophilicity with structure constraints. Overall, PepEVOLVE offers a practical, reproducible path to linear and macrocyclic peptide lead optimization when optimal edit sites are unknown, enabling more efficient chemical space exploration and improving design quality across multiple objectives to accelerate the design cycles.

Trieu Nguyen, Hao-Wei Pang, Shasha Feng

Merck & Co., Inc., Rahway, NJ, USA · University of South Florida, FL, USA

Molecular OptimizationDrug DiscoveryEvolutionary OptimizationReinforcement LearningMulti-Objective Evolutionary Optimization
AskPDF
Sep 29, 2025

Evolution Strategies at Scale: LLM Fine-Tuning Beyond Reinforcement Learning

Fine-tuning large language models (LLMs) for downstream tasks is an essential stage of modern AI deployment. Reinforcement learning (RL) has emerged as the dominant fine-tuning paradigm, underpinning many state-of-the-art LLMs. In contrast, evolution strategies (ES) has largely been overlooked due to the widespread belief that it does not scale to modern model sizes. This paper overturns this assumption by demonstrating the first successful application of ES to full-parameter fine-tuning of LLMs at the billion-parameter scale, without dimensionality reduction. ES can indeed search over extremely high-dimensional parameter spaces and outperform established RL implementations across multiple axes, including improved tolerance to long-horizon and delayed rewards, robustness across diverse base LLMs, reduced susceptibility to reward hacking, and improved training stability. These findings suggest that ES is not merely a viable alternative to RL, but a fundamentally different and powerful backpropagation-free post-training paradigm that opens a new direction for LLM fine-tuning beyond current RL-based approaches.

Xin Qiu, Yulu Gan, Conor F. Hayes +6

Cognizant AI Lab, San Francisco, CA, USA · Massachusetts Institute of Technology, Cambridge, MA, USA · University of California, Los Angeles, Los Angeles, CA, USA +1

Fine-TuningEvolutionary OptimizationLLM Fine-TuningEvolution StrategiesRL Fine-Tuning
AskPDF
Sep 10, 2025

A Systematic Survey on Large Language Models for Evolutionary Optimization: From Modeling to Solving

Large language models (LLMs) are increasingly integrated with evolutionary computation to support optimization tasks. This survey primarily focuses on evolutionary optimization, i.e., optimization based on evolutionary computation. For brevity, we use the term optimization throughout to denote this scope. However, existing surveys typically examine isolated roles of LLMs and do not provide a unified view that connects optimization modeling with optimization solving. To address this gap, we systematically review recent developments through a workflow-oriented framework. First, we organize the literature into two primary stages: LLMs for optimization modeling and LLMs for optimization solving (in this survey, the terms optimization modeling and optimization solving are used as concise forms of optimization problem modeling and optimization problem solving, respectively). Second, we divide the solving stage into three paradigms according to the role of the LLM: stand-alone optimizers, low-level components embedded in optimization algorithms, and high-level managers for algorithm selection and generation. Third, we analyze representative methods, identify their technical limitations, and clarify their relationships with traditional optimization approaches. We further substantiate this taxonomy through benchmark systematization, baseline comparisons, and practitioner-oriented guidance, and we review interdisciplinary applications across the natural sciences, engineering, and machine learning. Based on the resulting analysis, we identify research directions toward dynamic, self-evolving, and agentic optimization ecosystems. An up-to-date collection of related literature is maintained at https://github.com/ishmael233/LLM4OPT.

Yisong Zhang, Ran Cheng, Guoxing Yi +1

Evolutionary OptimizationLarge Language Model-Guided OptimizationLLMs for Evolutionary OptimizationNatural Language to Optimization Modeling
AskPDF
Jul 6, 2025

ESSA: Evolutionary Strategies for Scalable Alignment

Online alignment of large language models (LLMs) is dominated by reinforcement learning from human feedback (RLHF) with gradient-based optimizers such as PPO or GRPO. While effective, these pipelines require backpropagation through long rollouts, gradient synchronization across devices, and careful hyperparameter tuning, all of which become increasingly costly at scale. We present ESSA (Evolutionary Strategies for Scalable Alignment), a gradient-free online alignment stage that follows supervised fine-tuning (SFT) and replaces the gradient loop with inference-only black-box search. ESSA optimizes only the singular values of low-rank adaptation (LoRA) factors after a short SFT warm-start, restricting the search to a compact, task-aligned subspace where evolutionary search is practical even for 72B-parameter models. Because the loop is inference-only, ESSA is compatible with INT4/INT8 weight quantization and reduces inter-GPU communication to a few bytes per iteration. Across instruction following (IFEval), preference-based assistant tuning (HelpSteer2, HH-RLHF), and mathematical reasoning (GSM8K, MATH500), ESSA matches or exceeds LoRA-GRPO in the reported LoRA comparisons; on GSM8K it also outperforms Online DPO and PPO, while remaining competitive with both methods on IFEval. At scale, ESSA reaches a fixed MATH500 accuracy on Qwen2.5-32B/PRM800K up to 7.8x faster than LoRA-GRPO on 128 GPUs.

Daria Korotyshova, Boris Shaposhnikov, Alexey Malakhov +7

T-Tech · Yandex

LLM AlignmentEvolutionary OptimizationEvolution Strategies
AskPDF
Jul 2, 2025

Tuning without Peeking: Provable Generalization Bounds and Robust LLM Post-Training

Gradient-based optimization is the workhorse of deep learning, offering efficient and scalable training via backpropagation. However, exposing gradients during training can leak sensitive information about the underlying data, raising privacy and security concerns such as susceptibility to data poisoning attacks. In contrast, black-box optimization methods, which treat the model as an opaque function, relying solely on function evaluations to guide optimization, offer a promising alternative in scenarios where data access is restricted, adversarial risks are high, or overfitting is a concern. This paper introduces BBoxER, an evolutionary black-box method for LLM post-training that induces an information bottleneck via implicit compression of the training data. Leveraging the tractability of information flow, we provide non-vacuous generalization bounds and strong theoretical guarantees for robustness to data poisoning attacks and extraction attacks, while ensuring privacy by design. In experiments with LLMs, we demonstrate empirically that black-box optimization methods-despite the scalability and computational challenges inherent to black-box approaches-are able to learn, showing how a few iterations of BBoxER improve performance, generalize well on a benchmark of reasoning datasets, and are robust to membership inference attacks. This positions BBoxER as an attractive add-on on top of gradient-based optimization, offering suitability for deployment in restricted environments while also providing non-vacuous generalization guarantees.

Ismail Labiad, Mathurin Videau, Matthieu Kowalski +4

Meta FAIR · Université Paris-Saclay · LISN, Inria, CNRS +2

Privacy-Preserving Language ModelsEvolutionary OptimizationBlack-Box OptimizationInformation-Theoretic Generalization BoundsLLM Post-TrainingData Poisoning Attacks+1
AskPDF
Jun 30, 2025

Consensus-based optimization for closed-box adversarial attacks and a connection to evolution strategies

Consensus-based optimization (CBO) has established itself as an efficient gradient-free optimization scheme, with attractive mathematical properties, such as mean-field convergence results for non-convex loss functions. In this work, we study CBO in the context of closed-box adversarial attacks, which are imperceptible input perturbations that aim to fool a classifier, without accessing its gradient. Our contribution is to establish a connection between the so-called consensus hopping as introduced by Riedl et al. and natural evolution strategies (NES) commonly applied in the context of adversarial attacks and to rigorously relate both methods to gradient-based optimization schemes. Beyond that, we provide a comprehensive experimental study that shows that despite the conceptual similarities, CBO can outperform NES and other evolutionary strategies in certain scenarios.

Tim Roith, Leon Bungert, Philipp Wacker

TUM School of Computation, Information and Technology and Munich Center for Machine Learning (MCML), Boltzmannstr. 3, 85748 Garching bei München, Germany · Institute of Mathematics, Center for Artificial Intelligence and Data Science (CAIDAS), University of Würzburg, Emil-Fischer-Str. 40, 97074 Würzburg, Germany · School of Mathematics and Statistics, University of Canterbury

Evolutionary OptimizationBlack-Box Adversarial AttacksDerivative-Free Optimization
AskPDF
Jun 6, 2025

Decomposability-Guided Cooperative Coevolution for Large-Scale Itinerary Planning

Large-scale itinerary planning is a variant of the traveling salesman problem, aiming to determine an optimal path that maximizes the collected point-of-interest (POI) scores while minimizing travel time and cost, subject to travel duration constraints. This paper analyzes inter-component dependence in large-scale itinerary planning and establishes sufficient conditions for weak decomposability, under which an optimal path can visit each cluster as a contiguous subpath. Guided by this weak decomposability, we propose a novel multi-objective cooperative coevolutionary algorithm for large-scale itinerary planning, addressing the challenges of component imbalance and interactions. Specifically, we design a dynamic decomposition strategy based on the normalized fitness within each component, define optimization potential considering component scale and contribution, and develop a computational resource allocation strategy. Finally, we evaluate the proposed algorithm on 18 real-world test cases, eight large-scale semi-synthetic test cases, and eight adapted public clustered orienteering problem benchmark cases. Comparative results show that the proposed algorithm achieves the highest mean hypervolume across all 34 test cases and significantly outperforms the compared algorithms. Ablation and optimizer-decoupling analyses further validate the effectiveness of its core mechanisms and the general applicability of the proposed framework, with particularly pronounced performance gains on large-scale benchmark instances.

Ziyu Zhang, Peilan Xu, Yuetong Sun +2

School of Artificial Intelligence, Nanjing University of Information Science and Technology, Nanjing 210044, China. · Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen 518055, China. · School of Computer Science and Technology, Harbin Institute of Technology (Shenzhen), Shenzhen 518055, Guangdong, China. +1

Travel PlanningTraveling Salesperson ProblemEvolutionary OptimizationMulti-Objective Evolutionary Optimization
AskPDF
Mar 11, 2025

(θl,θuθ_l, θ_u)-Parametric Multi-Task Optimization: Joint Search in Solution and Infinite Task Spaces

Multi-task optimization is typically characterized by a fixed and finite set of tasks. The present paper relaxes this condition by considering a non-fixed and potentially infinite set of optimization tasks defined in a parameterized, continuous and bounded task space. We refer to this unique problem setting as parametric multi-task optimization (PMTO). Assuming the bounds of the task parameters to be (θl\boldsymbolθ_l, θu\boldsymbolθ_u), a novel (θl\boldsymbolθ_l, θu\boldsymbolθ_u)-PMTO algorithm is crafted to operate in two complementary modes. In an offline optimization mode, a joint search over solution and task spaces is carried out with the creation of two approximation models: (1) for mapping points in a unified solution space to the objective spaces of all tasks, which provably accelerates convergence by acting as a conduit for inter-task knowledge transfers, and (2) for probabilistically mapping tasks to their corresponding solutions, which facilitates evolutionary exploration of under-explored regions of the task space. In the online mode, the derived models enable direct optimization of any task within the bounds without the need to search from scratch. This outcome is validated on both synthetic test problems and practical case studies, with the significant real-world applicability of PMTO shown towards fast reconfiguration of robot controllers under changing task conditions. The potential of PMTO to vastly speedup the search for solutions to minimax optimization problems is also demonstrated through an example in robust engineering design.

Tingyang Wei, Jiao Liu, Abhishek Gupta +2

College of Computing and Data Science, Nanyang Technological University, Singapore · School of Mechanical Sciences, Indian Institute of Technology, Goa, India · Singapore Institute of Manufacturing Technology (SIMTech), Agency for Science, Technology and Research, Singapore +1

Evolutionary OptimizationMulti-Task Optimization
AskPDF
Jul 26, 2024

Small Molecule Optimization with Large Language Models

Molecular optimization, the process of designing molecules with desirable properties, represents a critical challenge in drug discovery. The recent advancements in large language models (LLMs) have opened new opportunities for their integration with traditional molecular optimization algorithms to improve performance. In this work, we propose Molecular Language Model powered Evolutionary Algorithm (Mol-E), an evolutionary algorithm that relies on the generative capabilities of LLMs trained on molecules and molecular properties. Scientific Contribution. Mol-E obtains the highest aggregate Top-10 AUC among the comparable full-23-task results considered here, scoring 17.500 in the task-agnostic regime, in which the oracle is treated strictly as a black box, and 20.551 in the task-informed regime, in which the optimizer receives a fixed semantic description of the objective. Mol-E also improves over the evaluated baselines on multi-property optimization with docking against DRD2, MK2, and AChE.

Philipp Guevorguian, Menua Bedrosian, Tigran Fahradyan +3

YerevaNN research lab, Yerevan, Armenia.

Molecular OptimizationEvolutionary OptimizationLarge Language Model-Guided OptimizationAI-Assisted Scientific ResearchMolecular GenerationDrug Discovery
AskPDF