mlsanityData to 14 Oct 2026

Neural Genetic Search in Discrete Spaces

arXiv 2502.10433 ↗First posted 9 Feb 2025cs.NE cs.LGcited by 5

Not on the map. Why?

Abstract (arXiv)

Effective search methods are crucial for improving the performance of deep generative models at test time. In this paper, we introduce a novel test-time search method, Neural Genetic Search (NGS), which incorporates the evolutionary mechanism of genetic algorithms into the generation procedure of deep models. The core idea behind NGS is its crossover, which is defined as parent-conditioned generation using trained generative models. This approach offers a versatile and easy-to-implement search algorithm for deep generative models. We demonstrate the effectiveness and flexibility of NGS through experiments across three distinct domains: routing problems, adversarial prompt generation for language models, and molecular design.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Builds on 48 of its 76 references that are in the map. First posted on arXiv 9 Feb 2025.

Builds on

48 of 76 references.

References · 48

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 3,442
  3. 2022Constitutional AI: Harmlessness from AI Feedbackcited by 3,050
  4. 2019The Curious Case of Neural Text Degenerationcited by 2,916
  5. 2024Gemma 2: Improving Open Language Models at a Practical Sizecited by 1,998
  6. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameterscited by 1,986
  7. 2022Red Teaming Language Models with Language Modelscited by 1,026
  8. 2024Large Language Monkeys: Scaling Inference Compute with Repeated Samplingcited by 991
  9. 2023AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language Modelscited by 785
  10. 2024Phi-4 Technical Reportcited by 711
  11. 2018Attention, Learn to Solve Routing Problems!cited by 656
  12. 2016Neural Combinatorial Optimization with Reinforcement Learningcited by 608
36 more
  1. 2021Flow Network based Generative Models for Non-Iterative Diverse Candidate Generationcited by 392
  2. 2018GuacaMol: Benchmarking Models for De Novo Molecular Designcited by 297
  3. 2021GFlowNet Foundationscited by 290
  4. 2024ReEvo: Large Language Models as Hyper-Heuristics with Reflective Evolutioncited by 276
  5. 2017Molecular De Novo Design through Deep Reinforcement Learningcited by 267
  6. 2020MiniLMv2: Multi-Head Self-Attention Relation Distillation for Compressing Pretrained Transformerscited by 245
  7. 2022Sample Efficiency Matters: A Benchmark for Practical Molecular Optimizationcited by 155
  8. 2023Neural Combinatorial Optimization with Heavy Decoder: Toward Large Scale Generalizationcited by 142
  9. 2022Evolution through Large Modelscited by 134
  10. 2023EvoPrompting: Language Models for Code-Level Neural Architecture Searchcited by 131
  11. 2022DIMES: A Differentiable Meta Solver for Combinatorial Optimization Problemscited by 129
  12. 2020Hybrid Genetic Search for the CVRP: Open-Source Implementation and SWAP* Neighborhoodcited by 125
  13. 2020Sample-Efficient Optimization in the Latent Space of Deep Generative Models via Weighted Retrainingcited by 123
  14. 2023Language Model Crossover: Variation through Few-Shot Promptingcited by 122
  15. 2022GFlowNets and variational inferencecited by 106
  16. 2021Efficient Active Search for Combinatorial Optimization Problemscited by 100
  17. 2021Deep Policy Dynamic Programming for Vehicle Routing Problemscited by 94
  18. 2021Conservative Objective Models for Effective Offline Model-Based Optimizationcited by 88
  19. 2022Simulation-guided Beam Search for Neural Combinatorial Optimizationcited by 84
  20. 2023DeepACO: Neural-enhanced Ant Systems for Combinatorial Optimizationcited by 78
  21. 2024Efficient Evolutionary Search Over Chemical Space with Large Language Modelscited by 61
  22. 2024Learning diverse attacks on large language models for robust red-teaming and safety tuningcited by 54
  23. 2024Ant Colony Sampling with GFlowNets for Combinatorial Optimizationcited by 53
  24. 2023PyVRP: a high-performance VRP solver packagecited by 52
  25. 2020Guiding Deep Molecular Optimization with Genetic Explorationcited by 51
  26. 2022Reinforced Genetic Algorithm for Structure-based Drug Designcited by 46
  27. 2021Amortized Tree Generation for Bottom-up Synthesis Planning and Synthesizable Molecular Designcited by 44
  28. 2021RoMA: Robust Model Adaptation for Offline Model-based Optimizationcited by 39
  29. 2024Genetic-guided GFlowNets for Sample Efficient Molecular Optimizationcited by 39
  30. 2022Bidirectional Learning for Offline Infinite-width Model-based Optimizationcited by 35
  31. 2023Meta-SAGE: Scale Meta-Learning Scheduled Adaptation with Guided Exploration for Mitigating Scale Shift on Combinatorial Optimizationcited by 23
  32. 2023Bootstrapped Training of Score-Conditioned Generator for Offline Design of Biological Sequencescited by 23
  33. 2024Molecule Generation with Fragment Retrieval Augmentationcited by 23
  34. 2023Bidirectional Learning for Offline Model-based Biological Sequence Designcited by 20
  35. 2024Generative Artificial Intelligence for Navigating Synthesizable Chemical Spacecited by 18
  36. 2024A Survey on LLM Inference-Time Self-Improvementcited by 17

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 3; 2026: 2.

5 papers.

  1. 2025A Genetic Algorithm for Navigating Synthesizable Molecular Spacescited by 4
  2. 2025USPR: Learning a Unified Solver for Profiled Routingcited by 1
  3. 2026Learning with Foresight: Enhancing Neural Routing Policy via Multi-Node Lookahead Predictioncited by 1
  4. 2026HyCO: A Hybrid Neural Solver for Combinatorial Optimizationcited by 1
  5. 2025Test-Time Search in Neural Graph Coarsening Procedures for the Capacitated Vehicle Routing Problemcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Deep Learning-Based Operators for Evolutionary Algorithms
  2. 2024Deep Neural Crossover
  3. 2021Evolutionary Generative Adversarial Networks with Crossover Based Knowledge Distillation
  4. 2021IE-GAN: An Improved Evolutionary Generative Adversarial Network Using a New Fitness Function and a Generic Crossover Operator
  5. 2025Evolution Meets Diffusion: Efficient Neural Architecture Generation
  6. 2025Attributes as Textual Genes: Leveraging LLMs as Genetic Algorithm Simulators for Conditional Synthetic Data Generation
  7. 2026Self-Evolving Deep Research via Joint Generation and Evaluation
  8. 2025Scaling Image and Video Generation via Test-Time Evolutionary Search