# What to Retrieve for Effective Retrieval-Augmented Code Generation? An Empirical Study and Beyond

arXiv: https://arxiv.org/abs/2503.20589 · HTML: https://mlsanity.com/paper/2503.20589/

## Key facts

- Cited by 22 papers in the map (data through 14 Oct 2026).
- Placed in RAG for code, one of 15 directions in Retrieval-augmented generation.
- Builds on 42 of its 61 references that are in the map.
- First posted on arXiv 26 Mar 2025.

## Placed in

- Retrieval-augmented generation / [RAG for code](https://mlsanity.com/rag/rag-for-code/)

## Builds on (most cited first, 30 of 42)

- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Qwen2.5 Technical Report](https://mlsanity.com/paper/2412.15115/) (2024)
- [Code Llama: Open Foundation Models for Code](https://mlsanity.com/paper/2308.12950/) (2023)
- [Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection](https://mlsanity.com/paper/2310.11511/) (2023)
- [Competition-Level Code Generation with AlphaCode](https://mlsanity.com/paper/2203.07814/) (2022)
- [DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligence](https://mlsanity.com/paper/2401.14196/) (2024)
- [Generalization through Memorization: Nearest Neighbor Language Models](https://mlsanity.com/paper/1911.00172/) (2019)
- [Making Retrieval-Augmented Language Models Robust to Irrelevant Context](https://mlsanity.com/paper/2310.01558/) (2023)
- [RepoCoder: Repository-Level Code Completion Through Iterative Retrieval and Generation](https://mlsanity.com/paper/2303.12570/) (2023)
- [UniXcoder: Unified Cross-Modal Pre-training for Code Representation](https://mlsanity.com/paper/2203.03850/) (2022)
- [RepoBench: Benchmarking Repository-Level Code Auto-Completion Systems](https://mlsanity.com/paper/2306.03091/) (2023)
- [CodeAgent: Enhancing Code Generation with Tool-Integrated Agent Systems for Real-World Repo-level Coding Challenges](https://mlsanity.com/paper/2401.07339/) (2024)
- [IntelliCode Compose: Code Generation Using Transformer](https://mlsanity.com/paper/2005.08025/) (2020)
- [CrossCodeEval: A Diverse and Multilingual Benchmark for Cross-File Code Completion](https://mlsanity.com/paper/2310.11248/) (2023)
- [DocPrompting: Generating Code by Retrieving the Docs](https://mlsanity.com/paper/2207.05987/) (2022)
- [Retrieval Augmented Code Generation and Summarization](https://mlsanity.com/paper/2108.11601/) (2021)
- [CodePlan: Repository-level Coding using LLMs and Planning](https://mlsanity.com/paper/2309.12499/) (2023)
- [ReACC: A Retrieval-Augmented Code Completion Framework](https://mlsanity.com/paper/2203.07722/) (2022)
- [CodeBERTScore: Evaluating Code Generation with Pretrained Models of Code](https://mlsanity.com/paper/2302.05527/) (2023)
- [RepoGraph: Enhancing AI Software Engineering with Repository-level Code Graph](https://mlsanity.com/paper/2410.14684/) (2024)
- [CERT: Continual Pre-Training on Sketches for Library-Oriented Code Generation](https://mlsanity.com/paper/2206.06888/) (2022)
- [ToolGen: Unified Tool Retrieval and Calling via Generation](https://mlsanity.com/paper/2410.03439/) (2024)
- [CoCoMIC: Code Completion By Jointly Modeling In-file and Cross-file Context](https://mlsanity.com/paper/2212.10007/) (2022)
- [Repoformer: Selective Retrieval for Repository-Level Code Completion](https://mlsanity.com/paper/2403.10059/) (2024)
- [CodeFill: Multi-token Code Completion by Jointly Learning from Structure and Naming Sequences](https://mlsanity.com/paper/2202.06689/) (2022)
- [Iterative Refinement of Project-Level Code Context for Precise Code Generation with Compiler Feedback](https://mlsanity.com/paper/2403.16792/) (2024)
- [Dataflow-Guided Retrieval Augmentation for Repository-Level Code Completion](https://mlsanity.com/paper/2405.19782/) (2024)
- [GraphCoder: Enhancing Repository-Level Code Completion via Code Context Graph-based Retrieval and Language Model](https://mlsanity.com/paper/2406.07003/) (2024)
- [Hot or Cold? Adaptive Temperature Sampling for Code Generation with Large Language Models](https://mlsanity.com/paper/2309.02772/) (2023)
- [REPOFUSE: Repository-Level Code Completion with Fused Dual Context](https://mlsanity.com/paper/2402.14323/) (2024)

## Built on by (most cited first, 22 of 22)

- [Retrieval-Augmented Code Generation: A Survey with Focus on Repository-Level Approaches](https://mlsanity.com/paper/2510.04905/) (2025)
- [An Empirical Study of Agent Developer Practices in AI Agent Frameworks](https://mlsanity.com/paper/2512.01939/) (2025)
- [RESCUE: Retrieval Augmented Secure Code Generation](https://mlsanity.com/paper/2510.18204/) (2025)
- [AlignCoder: Aligning Retrieval with Target Intent for Repository-Level Code Completion](https://mlsanity.com/paper/2601.19697/) (2026)
- [Context Engineering for Multi-Agent LLM Code Assistants Using Elicit, NotebookLM, ChatGPT, and Claude Code](https://mlsanity.com/paper/2508.08322/) (2025)
- [EffiReasonTrans: RL-Optimized Reasoning for Code Translation](https://mlsanity.com/paper/2510.18863/) (2025)
- [VeriCache: Turning Lossy KV Cache into Lossless LLM Inference](https://mlsanity.com/paper/2605.17613/) (2026)
- [RepoReasoner: Evaluating Repository-Level Code Reasoning Ability of Long-Context Language Models](https://mlsanity.com/paper/2607.25996/) (2026)
- [DRAINCODE: Stealthy Energy Consumption Attacks on Retrieval-Augmented Code Generation via Context Poisoning](https://mlsanity.com/paper/2601.20615/) (2026)
- [PhoenixRepair: Rethinking Repair Strategy Exploration in Software Agents](https://mlsanity.com/paper/2607.18859/) (2026)
- [LLMs as Packagers of HPC Software](https://mlsanity.com/paper/2511.05626/) (2025)
- [Bridging Generation and Training: A Systematic Review of Quality Issues in LLMs for Code](https://mlsanity.com/paper/2605.05267/) (2026)
- [Not All RAGs Are Created Equal: A Component-Wise Empirical Study for Software Engineering Tasks](https://mlsanity.com/paper/2605.14503/) (2026)
- [Beyond "What to Retrieve": Uncertainty in Retrieval-Augmented Code Generation](https://mlsanity.com/paper/2607.24884/) (2026)
- [Effective and Efficient Context Retrieval via Partial Dependency Graph for Repository-Level Code Generation](https://mlsanity.com/paper/2608.01927/) (2026)
- [SWE-Gate: Passing Functional Tests Is Not Enough for Software Engineering Agents](https://mlsanity.com/paper/2609.04167/) (2026)
- [Context-Augmented Code Generation Using Programming Knowledge Graphs](https://mlsanity.com/paper/2601.20810/) (2026)
- [LiveFMBench: Unveiling the Power and Limits of Agentic Workflows in Specification Generation](https://mlsanity.com/paper/2605.01394/) (2026)
- [ProjAgent: Procedural Similarity Retrieval for Repository-Level Code Generation](https://mlsanity.com/paper/2607.08691/) (2026)
- [Adaptive Critical Token-Aware Retrieval for Repository-Level Code Generation](https://mlsanity.com/paper/2609.01601/) (2026)
- [Efficient SWE Agent Benchmarking via Trajectory-Aware Evaluation](https://mlsanity.com/paper/2609.01603/) (2026)
- [MAP4CS: A Multi-dimensional Data Pruning Framework for Efficient Code Retriever Fine-tuning](https://mlsanity.com/paper/2610.11727/) (2026)

## Abstract (arXiv)

Repository-level code generation remains challenging due to complex code dependencies and the limitations of large language models (LLMs) in processing long contexts. While retrieval-augmented generation (RAG) frameworks are widely adopted, the effectiveness of different retrieved information sources-contextual code, APIs, and similar snippets-has not been rigorously analyzed. Through an empirical study on two benchmarks, we demonstrate that in-context code and potential API information significantly enhance LLM performance, whereas retrieved similar code often introduces noise, degrading results by up to 15%. Based on the preliminary results, we propose AllianceCoder, a novel context-integrated method that employs chain-of-thought prompting to decompose user queries into implementation steps and retrieves APIs via semantic description matching. Through extensive experiments on CoderEval and RepoExec, AllianceCoder achieves state-of-the-art performance, improving Pass@1 by up to 20% over existing approaches.