Beyond Single LLMs: Enhanced Code Generation via Multi-Stage Performance-Guided LLM Orchestration

• URL: http://arxiv.org/abs/2510.01379v1
• Date: Wed, 01 Oct 2025 19:07:16 GMT
• Title: Beyond Single LLMs: Enhanced Code Generation via Multi-Stage Performance-Guided LLM Orchestration
• Authors: Huashan Chen, Zhenyu Qi, Haotang Li, Hong Chen, Jinfu Chen, Kebin Peng, In Kee Kim, Kyu Hyung Lee, Sen He,
• Abstract summary: Large Language Models (LLMs) have become the predominant paradigm for automated code generation.<n>This paper challenges the single-model convention by introducing a multi-stage, performance-guided orchestration framework.<n>Perch orchestrates top-performing LLMs for each task context through stage-wise validation and rollback mechanisms.
• Score: 12.674888937998086
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: While Large Language Models (LLMs) have become the predominant paradigm for automated code generation, current single-model approaches fundamentally ignore the heterogeneous computational strengths that different models exhibit across programming languages, algorithmic domains, and development stages. This paper challenges the single-model convention by introducing a multi-stage, performance-guided orchestration framework that dynamically routes coding tasks to the most suitable LLMs within a structured generate-fix-refine workflow. Our approach is grounded in a comprehensive empirical study of 17 state-of-the-art LLMs across five programming languages (Python, Java, C++, Go, and Rust) using HumanEval-X benchmark. The study, which evaluates both functional correctness and runtime performance metrics (execution time, mean/max memory utilization, and CPU efficiency), reveals pronounced performance heterogeneity by language, development stage, and problem category. Guided by these empirical insights, we present PerfOrch, an LLM agent that orchestrates top-performing LLMs for each task context through stage-wise validation and rollback mechanisms. Without requiring model fine-tuning, PerfOrch achieves substantial improvements over strong single-model baselines: average correctness rates of 96.22% and 91.37% on HumanEval-X and EffiBench-X respectively, surpassing GPT-4o's 78.66% and 49.11%. Beyond correctness gains, the framework delivers consistent performance optimizations, improving execution time for 58.76% of problems with median speedups ranging from 17.67% to 27.66% across languages on two benchmarks. The framework's plug-and-play architecture ensures practical scalability, allowing new LLMs to be profiled and integrated seamlessly, thereby offering a paradigm for production-grade automated software engineering that adapts to the rapidly evolving generative AI landscape.

Related papers

• OPT-Engine: Benchmarking the Limits of LLMs in Optimization Modeling via Complexity Scaling [13.57588221678224]
  Large Language Models (LLMs) have demonstrated impressive progress in optimization modeling.<n>The boundaries of their capabilities in automated formulation and problem solving remain poorly understood.<n>We propose OPT-ENGINE, a benchmark framework designed to evaluate LLMs on optimization modeling with controllable and scalable difficulty levels.
  arXiv  Detail & Related papers  (2026-01-09T09:22:33Z)
• DiRL: An Efficient Post-Training Framework for Diffusion Language Models [54.405206032785706]
  Diffusion Language Models (dLLMs) have emerged as promising alternatives to Auto-Regressive (AR) models.<n>Existing methods suffer from computational inefficiency and objective mismatches between training and inference.<n>We introduce DiRL, an efficient post-training framework that tightly integrates FlexAttention-accelerated blockwise training with LMDeploy-optimized inference.
  arXiv  Detail & Related papers  (2025-12-23T08:33:19Z)
• LM4Opt-RA: A Multi-Candidate LLM Framework with Structured Ranking for Automating Network Resource Allocation [0.7933039558471408]
  We tackle complex analytical and mathematical reasoning tasks requiring nuanced contextual understanding.<n>Existing benchmarks datasets cannot address the complexities of such problems with dynamic environments, variables, and heterogeneous constraints.<n>We introduce NL4RA, a curated dataset comprising 50 resource allocation optimization problems formulated as LP, ILP, and MILP.<n>We then evaluate the performance of well-known open-source LLMs with varying parameter counts.
  arXiv  Detail & Related papers  (2025-11-13T23:19:43Z)
• MM-HELIX: Boosting Multimodal Long-Chain Reflective Reasoning with Holistic Platform and Adaptive Hybrid Policy Optimization [103.74675519953898]
  Long-chain reflective reasoning is a prerequisite for solving complex real-world problems.<n>We build a benchmark consisting 1,260 samples of 42 challenging synthetic tasks.<n>We generate post-training data and explore learning paradigms for exploiting such data.
  arXiv  Detail & Related papers  (2025-10-09T17:53:58Z)
• Exploring the Power of Diffusion Large Language Models for Software Engineering: An Empirical Investigation [27.11701612946034]
  DiffusionDLLMs offer a promising alternative to Autoregressive Large Language Models (AR-LLMs)<n>7BDLLMs outperform AR-LLMs with a 30% average accuracy improvement achieving a 113% gain on cross-file repair.
  arXiv  Detail & Related papers  (2025-10-06T09:13:25Z)
• TRACY: Benchmarking Execution Efficiency of LLM-Based Code Translation [15.302454413096335]
  We introduce TRACY, the first comprehensive benchmark designed to evaluate the execution efficiency of LLM-translated code.<n>The resulting benchmark comprises 1,011 code translation tasks across C++, Java, and Python.<n>Our work underscores the necessity of jointly optimizing for correctness and efficiency in future LLM-based code translation.
  arXiv  Detail & Related papers  (2025-08-15T13:33:52Z)
• CALM: Co-evolution of Algorithms and Language Model for Automatic Heuristic Design [11.639825726501659]
  Large language models (LLMs) can autonomously discover high-performings at a fraction of the traditional cost.<n>We propose a hybrid framework that combines verbal and numerical guidance.<n>Our method outperforms state-of-the-art (SOTA) baselines across various optimization tasks.
  arXiv  Detail & Related papers  (2025-05-18T07:48:47Z)
• TuRTLe: A Unified Evaluation of LLMs for RTL Generation [0.6010802600885173]
  We propose TuRTLe, a unified evaluation framework designed to assess LLMs across key RTL generation tasks.<n>We benchmark a diverse set of open LLMs and analyze their strengths and weaknesses in EDA-specific tasks.<n>Our results show that reasoning-based models, such as DeepSeek R1, consistently outperform others across multiple evaluation criteria.
  arXiv  Detail & Related papers  (2025-03-31T07:43:12Z)
• Collab: Controlled Decoding using Mixture of Agents for LLM Alignment [90.6117569025754]
  Reinforcement learning from human feedback has emerged as an effective technique to align Large Language models.<n>Controlled Decoding provides a mechanism for aligning a model at inference time without retraining.<n>We propose a mixture of agent-based decoding strategies leveraging the existing off-the-shelf aligned LLM policies.
  arXiv  Detail & Related papers  (2025-03-27T17:34:25Z)
• EquiBench: Benchmarking Large Language Models' Reasoning about Program Semantics via Equivalence Checking [58.15568681219339]
  We introduce EquiBench, a new benchmark for evaluating large language models (LLMs)<n>This task directly tests a model's ability to reason about program semantics.<n>We evaluate 19 state-of-the-art LLMs and find that in the most challenging categories, the best accuracies are 63.8% and 76.2%, only modestly above the 50% random baseline.
  arXiv  Detail & Related papers  (2025-02-18T02:54:25Z)
• Adaptive Pruning for Large Language Models with Structural Importance Awareness [66.2690963378878]
  Large language models (LLMs) have significantly improved language understanding and generation capabilities.<n>LLMs are difficult to deploy on resource-constrained edge devices due to their high computational and storage resource demands.<n>We propose structurally-aware adaptive pruning (SAAP) to significantly reduce the computational and memory costs while maintaining model performance.
  arXiv  Detail & Related papers  (2024-12-19T18:08:04Z)
• Bridging the Language Gaps in Large Language Models with Inference-Time Cross-Lingual Intervention [71.12193680015622]
  Large Language Models (LLMs) have shown remarkable capabilities in natural language processing. LLMs exhibit significant performance gaps among different languages. We propose Inference-Time Cross-Lingual Intervention (INCLINE) to overcome these limitations without incurring significant costs.
  arXiv  Detail & Related papers  (2024-10-16T11:23:03Z)
• Mono-InternVL: Pushing the Boundaries of Monolithic Multimodal Large Language Models with Endogenous Visual Pre-training [48.455597568212944]
  We present Mono-InternVL, a novel monolithic MLLM that seamlessly integrates a set of visual experts via a multimodal mixture-of-experts structure.<n>In particular, EViP is designed as a progressive learning process for visual experts, which aims to fully exploit the visual knowledge from noisy data to high-quality data.
  arXiv  Detail & Related papers  (2024-10-10T17:59:22Z)
• Accelerating LLaMA Inference by Enabling Intermediate Layer Decoding via Instruction Tuning with LITE [62.13435256279566]
  Large Language Models (LLMs) have achieved remarkable performance across a wide variety of natural language tasks. However, their large size makes their inference slow and computationally expensive. We show that it enables these layers to acquire 'good' generation ability without affecting the generation ability of the final layer.
  arXiv  Detail & Related papers  (2023-10-28T04:07:58Z)
• Learning Performance-Improving Code Edits [107.21538852090208]
  We introduce a framework for adapting large language models (LLMs) to high-level program optimization. First, we curate a dataset of performance-improving edits made by human programmers of over 77,000 competitive C++ programming submission pairs. For prompting, we propose retrieval-based few-shot prompting and chain-of-thought, and for finetuning, these include performance-conditioned generation and synthetic data augmentation based on self-play.
  arXiv  Detail & Related papers  (2023-02-15T18:59:21Z)

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.