C-Evolve: Consensus-based Evolution for Prompt Groups

• URL: http://arxiv.org/abs/2509.23331v1
• Date: Sat, 27 Sep 2025 14:35:49 GMT
• Title: C-Evolve: Consensus-based Evolution for Prompt Groups
• Abstract summary: We introduce Consensus-Evolve (C-Evolve), an evolutionary algorithm that discovers a group of prompts whose aggregated outputs achieve optimal performance.<n>C-Evolve employs an island-based evolutionary algorithm to maintain population diversity, and prompts from distinct islands are selected to form groups.<n>Our method achieves state-of-the-art performance across a wide range of tasks, including both open-ended tasks like HotpotQA and closed-ended tasks like MATH.
• Score: 43.256446610084986
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Prompt evolution algorithms offer a powerful paradigm for enhancing AI systems based on closed-source models, while few work explores whether aggregating results from multiple prompts to reach a consensus can further advance the system capability boundary. In this paper, we introduce Consensus-Evolve (C-Evolve), an evolutionary algorithm that discovers a group of prompts whose aggregated outputs after majority voting achieve optimal performance. More specifically, C-Evolve employs an island-based evolutionary algorithm to maintain population diversity, and prompts from distinct islands are selected to form groups to aggregate their outputs. The key difference from single individual evolution is a voting score, which evaluates each individual prompt's contribution within groups. We take this as the fitness score for evolution instead of individual performance. Consequently, C-Evolve is more likely to produce and maintain prompts with higher potential to form a high-performing group and eliminate low-performing ones, gradually improving the group performance after reaching consensus. Our method achieves state-of-the-art performance across a wide range of tasks, including both open-ended tasks like HotpotQA and closed-ended tasks like MATH. On Qwen3-8B, C-Evolve achieves 70.67% on HotpotQA and 43.88% on IFBench, which are 4.95% and 2.73% higher than GEPA, respectively. For GPT-4.1-mini, the accuracy on IFBench is further improved to 47.96% and reaches 95.33% in the MATH benchmark. These results demonstrate the C-Evolve's competitive performance.

Related papers

• SERPO: Self-Evolving Rubric Policy Optimization for Open-Ended Test-Time Reinforcement Learning [14.505575259468053]
  Test-time reinforcement learning (TTRL) enables language models to self-evolve at inference time without labeled feedback.<n>We introduce SERPO, which replaces answer voting with a closed loop that co-evolves response evidence, querytokens, and policy parameters.<n>New actor rollouts then refresh both the archives and rubrics, closing the three-way evolution loop.
  arXiv  Detail & Related papers  (2026-07-29T13:03:07Z)
• Teaching LLMs to Self-Evolve: Cultivating Core Meta-Skills with Reinforcement Learning [51.17262419591891]
  Test-time scaling through iterative self-evolution with environment feedback, as demonstrated by AlphaEvolve, shows remarkable performance gains.<n>We present MetaEvolve, a framework designed to develop these meta-skills via a data synthesis pipeline, evolution-aware reinforcement learning, and inference-time evolutionary search.
  arXiv  Detail & Related papers  (2026-07-24T04:35:29Z)
• EvoX: Meta-Evolution for Automated Discovery [115.89434419482797]
  EvoX is an adaptive evolution method that optimize its own evolution process.<n>It continuously updates how prior solutions are selected and varied based on progress.<n>It outperforms existing AI-driven evolutionary methods including AlphaEvolve, OpenEvolve, GEPA, and ShinkaEvolve on the majority of tasks.
  arXiv  Detail & Related papers  (2026-02-26T18:54:41Z)
• Group-Evolving Agents: Open-Ended Self-Improvement via Experience Sharing [26.102738613616456]
  Group-Evolving Agents (GEA) treats a group of agents as the fundamental evolutionary unit.<n>We evaluate GEA on challenging coding benchmarks, where it significantly outperforms state-of-the-art self-evolving methods.
  arXiv  Detail & Related papers  (2026-02-04T18:29:36Z)
• An Ensemble of Evolutionary Algorithms With Both Crisscross Search and Sparrow Search for Processing Inferior Individuals [13.79887031590359]
  We introduce two types of evolutionary algorithm proposed recently - crisscross search and sparrow search - into EA4eig as secondary evolutionary algorithms to process inferior individuals.<n> Experimental results show that our EA4eigCS outperforms EA4eig and is competitive when compared with state-of-the-art algorithms.
  arXiv  Detail & Related papers  (2026-01-15T10:36:08Z)
• Evolving Deep Learning Optimizers [0.0]
  We present a genetic algorithm framework for discovering deep learning optimization algorithms.<n>Our approach encodes genomes that specify combinations of primitive update terms.<n>We find an evolved that outperforms Adam by 2.6% in aggregate fitness.
  arXiv  Detail & Related papers  (2025-12-05T01:30:07Z)
• RoundTable: Investigating Group Decision-Making Mechanism in Multi-Agent Collaboration [49.4875652673051]
  We analyze how different voting rules affect decision quality and efficiency in a multi-round collaboration.<n>At the extreme, unanimous voting gives 87% lower initial performance than the best-performing method.<n>Our findings highlight the crucial role of group decision-making in optimizing MAS collaboration.
  arXiv  Detail & Related papers  (2024-11-11T17:37:47Z)
• Evolving a multi-population evolutionary-QAOA on distributed QPUs [0.0]
  We benchmark this Evolutionary-QAOA (E-QAOA) approach on the Max-Cut problem for $d$-3 regular graphs of 4 to 26 nodes.<n>We propose a novel distributed multi-population EA strategy, executing parallel, independent populations on two quantum processing units.
  arXiv  Detail & Related papers  (2024-09-16T21:16:51Z)
• ConvBench: A Comprehensive Benchmark for 2D Convolution Primitive Evaluation [0.34952465649465553]
  This paper proposes ConvBench, a primitive-level benchmark for the evaluation and comparison of convolution algorithms. It assesses 9243 convolution operations derived from 1097 real-world deep learning models. The experiments showed results faster than Im2col-GEMM in 93.6% of the convolutions.
  arXiv  Detail & Related papers  (2024-07-15T13:58:24Z)
• DARLEI: Deep Accelerated Reinforcement Learning with Evolutionary Intelligence [77.78795329701367]
  We present DARLEI, a framework that combines evolutionary algorithms with parallelized reinforcement learning. We characterize DARLEI's performance under various conditions, revealing factors impacting diversity of evolved morphologies. We hope to extend DARLEI in future work to include interactions between diverse morphologies in richer environments.
  arXiv  Detail & Related papers  (2023-12-08T16:51:10Z)
• Effective Mutation Rate Adaptation through Group Elite Selection [50.88204196504888]
  This paper introduces the Group Elite Selection of Mutation Rates (GESMR) algorithm. GESMR co-evolves a population of solutions and a population of MRs, such that each MR is assigned to a group of solutions. With the same number of function evaluations and with almost no overhead, GESMR converges faster and to better solutions than previous approaches.
  arXiv  Detail & Related papers  (2022-04-11T01:08:26Z)
• Quality Evolvability ES: Evolving Individuals With a Distribution of Well Performing and Diverse Offspring [0.2578242050187029]
  Evolvability algorithms aim to automatically learn good genetic representations. Evolvability ES allows direct selection for evolvability with little computation. We propose Quality Evolvability ES, a method that simultaneously optimize for task performance and evolvability.
  arXiv  Detail & Related papers  (2021-03-19T13:22:22Z)
• Involution: Inverting the Inherence of Convolution for Visual Recognition [72.88582255910835]
  We present a novel atomic operation for deep neural networks by inverting the principles of convolution, coined as involution. The proposed involution operator could be leveraged as fundamental bricks to build the new generation of neural networks for visual recognition. Our involution-based models improve the performance of convolutional baselines using ResNet-50 by up to 1.6% top-1 accuracy, 2.5% and 2.4% bounding box AP, and 4.7% mean IoU absolutely.
  arXiv  Detail & Related papers  (2021-03-10T18:40:46Z)
• Inception Convolution with Efficient Dilation Search [121.41030859447487]
  Dilation convolution is a critical mutant of standard convolution neural network to control effective receptive fields and handle large scale variance of objects. We propose a new mutant of dilated convolution, namely inception (dilated) convolution where the convolutions have independent dilation among different axes, channels and layers. We explore a practical method for fitting the complex inception convolution to the data, a simple while effective dilation search algorithm(EDO) based on statistical optimization is developed.
  arXiv  Detail & Related papers  (2020-12-25T14:58:35Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.