# Revisiting Code Search in a Two-Stage Paradigm

arXiv: https://arxiv.org/abs/2208.11274 · HTML: https://mlsanity.com/paper/2208.11274/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Aug 2022.

## Built on by (most cited first, 8 of 8)

- [Agents in Software Engineering: Survey, Landscape, and Vision](https://mlsanity.com/paper/2409.09030/) (2024)
- [AlignCoder: Aligning Retrieval with Target Intent for Repository-Level Code Completion](https://mlsanity.com/paper/2601.19697/) (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)
- [MediaWiki Code2Code Search: Neural Retrieval for the Semantic Discovery of Open-Source Software Entities](https://mlsanity.com/paper/2607.26766/) (2026)
- [Recall Before Rerank: Benchmarking Deep Learning Models for Large-Scale Code-to-Code Retrieval](https://mlsanity.com/paper/2606.27401/) (2026)
- [SynH-Rank: Quality-Aware Code Search via Diverse Data Synthesis and Hierarchical Ranking Training](https://mlsanity.com/paper/2607.17139/) (2026)
- [CodeGraph: Open-Taxonomy Knowledge Graph for Source Code with Wikidata Grounding](https://mlsanity.com/paper/2609.29474/) (2026)

## Abstract (arXiv)

With a good code search engine, developers can reuse existing code snippets and accelerate software development process. Current code search methods can be divided into two categories: traditional information retrieval (IR) based and deep learning (DL) based approaches. DL-based approaches include the cross-encoder paradigm and the bi-encoder paradigm. However, both approaches have certain limitations. The inference of IR-based and bi-encoder models are fast, however, they are not accurate enough; while cross-encoder models can achieve higher search accuracy but consume more time. In this work, we propose TOSS, a two-stage fusion code search framework that can combine the advantages of different code search methods. TOSS first uses IR-based and bi-encoder models to efficiently recall a small number of top-k code candidates, and then uses fine-grained cross-encoders for finer ranking. Furthermore, we conduct extensive experiments on different code candidate volumes and multiple programming languages to verify the effectiveness of TOSS. We also compare TOSS with six data fusion methods. Experimental results show that TOSS is not only efficient, but also achieves state-of-the-art accuracy with an overall mean reciprocal ranking (MRR) score of 0.763, compared to the best baseline result on the CodeSearchNet benchmark of 0.713. Our source code and experimental data are available at: https://github.com/fly-dragon211/TOSS.