# Enhancing LLM-Based Code Generation with Complexity Metrics: A Feedback-Driven Approach

arXiv: https://arxiv.org/abs/2505.23953 · HTML: https://mlsanity.com/paper/2505.23953/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 35 references that are in the map.
- First posted on arXiv 29 May 2025.

## Builds on (most cited first, 16 of 16)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [A Unified Approach to Interpreting Model Predictions](https://mlsanity.com/paper/1705.07874/) (2017)
- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [Program Synthesis with Large Language Models](https://mlsanity.com/paper/2108.07732/) (2021)
- [Code Llama: Open Foundation Models for Code](https://mlsanity.com/paper/2308.12950/) (2023)
- [BigCodeBench: Benchmarking Code Generation with Diverse Function Calls and Complex Instructions](https://mlsanity.com/paper/2406.15877/) (2024)
- [Debug like a Human: A Large Language Model Debugger via Verifying Runtime Execution Step-by-step](https://mlsanity.com/paper/2402.16906/) (2024)
- [Parsel: Algorithmic Reasoning with Language Models by Composing Decompositions](https://mlsanity.com/paper/2212.10561/) (2022)
- [ANPL: Towards Natural Programming with Interactive Decomposition](https://mlsanity.com/paper/2305.18498/) (2023)
- [No Man is an Island: Towards Fully Automatic Programming by Code Search, Code Generation and Program Repair](https://mlsanity.com/paper/2409.03267/) (2024)
- [A Pragmatic Approach for Hyper-Parameter Tuning in Search-based Test Case Generation](https://mlsanity.com/paper/2101.05738/) (2021)
- [An Empirical Study on Bug Severity Estimation using Source Code Metrics and Static Analysis](https://mlsanity.com/paper/2206.12927/) (2022)

## Built on by (most cited first, 10 of 10)

- [Prompt to Pwn: Automated Exploit Generation for Smart Contracts](https://mlsanity.com/paper/2508.01371/) (2025)
- [Code Is More Than Text: Uncertainty Estimation for Code Generation](https://mlsanity.com/paper/2606.09577/) (2026)
- [LogicIF: Towards Complex Logic Instruction Following](https://mlsanity.com/paper/2508.09125/) (2025)
- [Throttling Web Agents Using Reasoning Gates](https://mlsanity.com/paper/2509.01619/) (2025)
- [Not All Code Is Equal: A Data-Centric Study of Code Complexity and LLM Reasoning](https://mlsanity.com/paper/2601.21894/) (2026)
- [CVeDRL: An Efficient Code Verifier via Difficulty-aware Reinforcement Learning](https://mlsanity.com/paper/2601.22803/) (2026)
- [SwiftSolve: A Self-Iterative, Complexity-Aware Multi-Agent Framework for Competitive Programming](https://mlsanity.com/paper/2510.22626/) (2025)
- [Beyond Accuracy: Characterizing Code Comprehension Capabilities in (Large) Language Models](https://mlsanity.com/paper/2601.12951/) (2026)
- [How Generation Architecture Shapes Code Complexity in Multi-Agent LLM Systems: A Paired Study on HumanEval](https://mlsanity.com/paper/2606.00308/) (2026)
- [Complexity-Aware Evaluation of LLM Comprehension](https://mlsanity.com/paper/2609.37405/) (2026)

## Abstract (arXiv)

Automatic code generation has gained significant momentum with the advent of Large Language Models (LLMs) such as GPT-4. Although many studies focus on improving the effectiveness of LLMs for code generation, very limited work tries to understand the generated code's characteristics and leverage that to improve failed cases. In this paper, as the most straightforward characteristic of code, we investigate the relationship between code complexity and the success of LLM generated code. Using a large set of standard complexity metrics, we first conduct an empirical analysis to explore their correlation with LLM's performance on code generation (i.e., Pass@1). Using logistic regression models, we identify which complexity metrics are most predictive of code correctness. Building on these findings, we propose an iterative feedback method, where LLMs are prompted to generate correct code based on complexity metrics from previous failed outputs. We validate our approach across multiple benchmarks (i.e., HumanEval, MBPP, LeetCode, and BigCodeBench) and various LLMs (i.e., GPT-4o, GPT-3.5 Turbo, Llama 3.1, and GPT-o3 mini), comparing the results with two baseline methods: (a) zero-shot generation, and (b) iterative execution-based feedback without our code complexity insights. Experiment results show that our approach makes notable improvements, particularly with a smaller LLM (GPT3.5 Turbo), where, e.g., Pass@1 increased by 35.71% compared to the baseline's improvement of 12.5% on the HumanEval dataset. The study expands experiments to BigCodeBench and integrates the method with the Reflexion code generation agent, leading to Pass@1 improvements of 20% (GPT-4o) and 23.07% (GPT-o3 mini). The results highlight that complexity-aware feedback enhances both direct LLM prompting and agent-based workflows.