CoRNStack: High-Quality Contrastive Data for Better Code Ranking

By Tarun Suresh, Revanth Gangi Reddy, Yifei Xu, Zach Nussbaum, Andriy Mulyar, Brandon Duderstadt, Heng Ji. arXiv.org, 2024

12 citations recorded by the source corpus.

Subfields: Information Retrieval & Search, Representation & Self-Supervised Learning, Code Generation & Program Synthesis

View Paper: https://doi.org/10.48550/arXiv.2412.01007

Citations in the corpus

• Neurosymbolic Repo-level Code Localization (2026, unknown): “Following Suresh et al. [25], we retained 274 of 300 original instances where patches modify existing functions or classes.”
• Bagging-Based Model Merging for Robust General Text Embeddings (2026, unknown): “Furthermore, to support code retrieval capabilities, we incorporate code-specific training examples sourced from Suresh et al. [51].”
• SweRank+: Multilingual, Multi-Turn Code Ranking for Software Issue Localization (2025, unknown): “To improve efficiency, S WE R-ANK (Reddy et al., 2025) reformulates issue localization as a retrieve-and-rerank problem, leveraging specialized bi-encoder retrievers and LLM rerankers trained on curated datasets (Suresh et al., 2024).”
• Across Programming Language Silos: A Study on Cross-Lingual Retrieval-augmented Code Generation (2025, unknown): “For the retrieval phase, we employ a state-of-the-art embedding model CodeRankEmbed [46] for the code retrieval task, which supports multi-lingual code retrieval.”
• SweRank: Software Issue Localization with Code Ranking (2025, unknown): “Suresh et al. (2024) improve the contrastive training process between function snippets and associated doc-strings with better consistency filtering and harder negative mining.”
• FreshStack: Building Realistic Benchmarks for Evaluating Retrieval on Technical Documents (2025, unknown): “• Providing as much information as possible in supplemental material (appended to the paper) is recommended, but including URLs to data and code is permitted.”
• LocAgent: Graph-Guided LLM Agents for Code Localization (2025, unknown): “…state-of-the-art embedding models, including the general-purpose E5-base-v2 (Wang et al., 2022) and specialized code embedding models such as Jina-Code-v2 (Günther et al., 2023), Codesage-large-v2 (Zhang et al., 2024), and the current SOTA code embedding model CodeRankEmbed (Suresh et al., 2024).”
• CRANE: Reasoning with constrained LLM generation (2025, unknown): “…used in AI systems that interact with traditional software tools like Python interpreters (OpenAI, 2024; Chen et al., 2023) for code generation (Suresh et al., 2024a;b; Jiang et al., 2024), logical solvers (Pan et al., 2023; Olaus-* son et al., 2023), and theorem provers (Wu et al., 2022; Yang…”
• Is The Watermarking Of LLM-Generated Code Robust? (2024, unknown): “…like GPT and Codex in understanding and generating code holds transformative potential for software development (Chen et al., 2021; OpenAI, 2023; Suresh et al., 2024b; Reddy et al., 2025; Wei et al., 2025b;a;d;c), especially when combined with constrained decoding algorithms (Ugare et al.,…”