# Tackling Long Code Search with Splitting, Encoding, and Aggregating

arXiv: https://arxiv.org/abs/2208.11271 · HTML: https://mlsanity.com/paper/2208.11271/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Aug 2022.

## Built on by (most cited first, 9 of 9)

- [Agents in Software Engineering: Survey, Landscape, and Vision](https://mlsanity.com/paper/2409.09030/) (2024)
- [Beyond Functional Correctness: Investigating Coding Style Inconsistencies in Large Language Models](https://mlsanity.com/paper/2407.00456/) (2024)
- [An Empirical Study of Agent Developer Practices in AI Agent Frameworks](https://mlsanity.com/paper/2512.01939/) (2025)
- [AlignCoder: Aligning Retrieval with Target Intent for Repository-Level Code Completion](https://mlsanity.com/paper/2601.19697/) (2026)
- [EffiReasonTrans: RL-Optimized Reasoning for Code Translation](https://mlsanity.com/paper/2510.18863/) (2025)
- [Code Search Debiasing:Improve Search Results beyond Overall Ranking Performance](https://mlsanity.com/paper/2311.14901/) (2023)
- [ShortCoder: Knowledge-Augmented Syntax Optimization for Token-Efficient Code Generation](https://mlsanity.com/paper/2601.09703/) (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)

## Abstract (arXiv)

Code search with natural language helps us reuse existing code snippets. Thanks to the Transformer-based pretraining models, the performance of code search has been improved significantly. However, due to the quadratic complexity of multi-head self-attention, there is a limit on the input token length. For efficient training on standard GPUs like V100, existing pretrained code models, including GraphCodeBERT, CodeBERT, RoBERTa (code), take the first 256 tokens by default, which makes them unable to represent the complete information of long code that is greater than 256 tokens. To tackle the long code problem, we propose a new baseline SEA (Split, Encode and Aggregate), which splits long code into code blocks, encodes these blocks into embeddings, and aggregates them to obtain a comprehensive long code representation. With SEA, we could directly use Transformer-based pretraining models to model long code without changing their internal structure and re-pretraining. We also compare SEA with sparse Trasnformer methods. With GraphCodeBERT as the encoder, SEA achieves an overall mean reciprocal ranking score of 0.785, which is 10.1% higher than GraphCodeBERT on the CodeSearchNet benchmark, justifying SEA as a strong baseline for long code search. Our source code and experimental data are available at: https://github.com/fly-dragon211/SEA.