Pith. sign in

REVIEW 13 cited by

RepoFusion: Training Code Models to Understand Your Repository

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2306.10998 v1 pith:QA4BKTAB submitted 2023-06-19 cs.LG cs.AIcs.PLcs.SE

classification cs.LGcs.AIcs.PLcs.SE
keywords contextrepositorycodemodelslargerrepofusiontrainedtraining
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
abstract

Despite the huge success of Large Language Models (LLMs) in coding assistants like GitHub Copilot, these models struggle to understand the context present in the repository (e.g., imports, parent classes, files with similar names, etc.), thereby producing inaccurate code completions. This effect is more pronounced when using these assistants for repositories that the model has not seen during training, such as proprietary software or work-in-progress code projects. Recent work has shown the promise of using context from the repository during inference. In this work, we extend this idea and propose RepoFusion, a framework to train models to incorporate relevant repository context. Experiments on single-line code completion show that our models trained with repository context significantly outperform much larger code models as CodeGen-16B-multi ($\sim73\times$ larger) and closely match the performance of the $\sim 70\times$ larger StarCoderBase model that was trained with the Fill-in-the-Middle objective. We find these results to be a novel and compelling demonstration of the gains that training with repository context can bring. We carry out extensive ablation studies to investigate the impact of design choices such as context type, number of contexts, context length, and initialization within our framework. Lastly, we release Stack-Repo, a dataset of 200 Java repositories with permissive licenses and near-deduplicated files that are augmented with three types of repository contexts. Additionally, we are making available the code and trained checkpoints for our work. Our released resources can be found at \url{https://huggingface.co/RepoFusion}.

Discussion (0). Sign in to comment.

Forward citations

Cited by 13 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. In Line with Context: Repository-Level Code Generation via Context Inlining

    cs.SE 2026-01 unverdicted novelty 7.0 of 10

    InlineCoder reframes repository-level code generation as function-level coding by using a draft anchor to inline the target function into its call graph for upstream usage and downstream dependency context.

  2. SWE-QA: Can Language Models Answer Repository-level Code Questions?

    cs.CL 2025-09 unverdicted novelty 7.0 of 10

    SWE-QA creates a new repository-level code QA benchmark with 576 pairs and an agentic LLM framework, showing promise but open challenges for models handling complex codebases.

  3. Code2LoRA: Hypernetwork-Generated Adapters for Code Language Models under Software Evolution

    cs.SE 2026-06 unverdicted novelty 6.0 of 10

    Code2LoRA generates repo-specific LoRA adapters via hypernetwork for code LMs, matching per-repo LoRA on static tasks and exceeding shared LoRA by 5.2 pp on evolving code in a 604-repo benchmark.

  4. REAgent: Requirement-Driven LLM Agents for Software Issue Resolution

    cs.SE 2026-04 unverdicted novelty 6.0 of 10

    REAgent improves LLM patch generation for software issues by 17.4% on average through automated construction, quality checking, and iterative refinement of structured issue-oriented requirements.

  5. Better Call Grep: Evaluating and Improving Grep-Like Lexical Retrieval for Repository-Level Code Completion

    cs.SE 2026-01 conditional novelty 6.0 of 10

    LLM-generated ripgrep queries plus BM25 re-ranking and line-interval de-duplication outperform graph- and RL-based retrievers for repository-level code completion on CrossCodeEval and RepoEval-Updated.

  6. RepoDebug: Repository-Level Multi-Task and Multi-Language Debugging Evaluation of Large Language Models

    cs.SE 2025-09 conditional novelty 6.0 of 10

    RepoDebug is a new multi-language, multi-task benchmark for repository-level code debugging on which current LLMs, including the best model Claude 3.5 Sonnet, perform poorly.

  7. Exploring the Challenges and Opportunities of AI-assisted Codebase Generation

    cs.SE 2025-08 conditional novelty 6.0 of 10

    Developers prompting codebase-level AI assistants are often dissatisfied with generated code, citing missing functionality, poor code quality, and communication gaps, despite varied prompting strategies.

  8. Enhancing Project-Specific Code Completion by Inferring Internal API Information

    cs.SE 2025-07 conditional novelty 6.0 of 10

    A retrieval-augmented code completion method that infers project-internal APIs from a rough draft and a static knowledge base, beating existing repo-level baselines.

  9. Optimization Hyper-parameter Laws for Large Language Models

    cs.LG 2024-09 unverdicted novelty 6.0 of 10

    Opt-Laws predicts LLM final training loss from LR schedules via SDE-derived convergence and escape features, with 94% Top-2 hit rate on held-out schedules and F1=0.92 for divergence detection.

  10. Benchmarking LLM-Based Static Analysis for Secure Smart Contract Development: Reliability, Limitations, and Potential Hybrid Solutions

    cs.CR 2026-05 unverdicted novelty 5.0 of 10

    LLMs for smart contract security analysis show lexical bias from identifier names causing high false positives, with prompting creating precision-recall trade-offs, positioning them as complements rather than replacem...

  11. GRACE: Graph-Guided Repository-Aware Code Completion through Hierarchical Code Fusion

    cs.SE 2025-09 conditional novelty 5.0 of 10

    GRACE combines a multi-level code graph, hybrid text-structure retrieval, and graph fusion to improve repository-level code completion over vanilla and graph-based RAG baselines.

  12. Retrieval-Augmented Generation for AI-Generated Content: A Survey

    cs.CV 2024-02 accept novelty 5.0 of 10

    A survey classifying RAG foundations for AIGC, summarizing enhancements, cross-modal applications, benchmarks, limitations, and future directions.

  13. A Survey on Large Language Models for Code Generation

    cs.CL 2024-06 unverdicted novelty 3.0 of 10

    A systematic literature review that organizes recent work on LLMs for code generation into a taxonomy covering data curation, model advances, evaluations, ethics, environmental impact, and applications, with benchmark...

Pith tools