Research map / Large language models / Agents, code and math
CLARC: C/C++ Benchmark for Robust Code Search
Kaicheng Wang, Liyan Huang, Weike Fang, Weihang Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace identifiers with neutral names (neutralized setting) | Voyage-code-3 · CLARC Group 2 | MRR: 92.10 -> 77.18 | Hurt |
| randomize identifier names | Voyage-code-3 · CLARC Group 2 | MRR: 92.10 -> 69.43 | Hurt |
| neutralize identifiers | Nomic · CLARC Group 2 | MRR: 91.61 -> 67.65 | Hurt |
| neutralize identifiers | OpenAI-text-embedding-large · CLARC Group 2 | MRR: 81.66 -> 37.14 | Hurt |
| randomize identifiers | CodeT5+ · CLARC Group 1 | MRR: 58.84 -> 29.52 | Hurt |
| compile to Assembly instead of C/C++ | Voyage-code-3 · CLARC Group 1 | MRR: 86.93 -> 29.02 | Hurt |
| compile to Assembly instead of C/C++ | OpenAI-text-embedding-large · CLARC Group 1 | MRR: 80.16 -> 8.61 | Hurt |
| fine-tune on standard training set | CodeT5+ · CLARC Group 1 (standard) | MRR: 58.84 -> 76.51 | Improved |
| fine-tune on randomized training set | CodeT5+ · CLARC Group 1 (randomized) | MRR: 29.52 -> 60.93 | Improved |
| fine-tune on standard training set | OASIS · CLARC Group 2 (standard) | MRR: 88.30 -> 78.26 | Hurt |
| fine-tune on randomized training set | OASIS · CLARC Group 3 (randomized) | MRR: 30.51 -> 15.53 | Hurt |
| fine-tune on neutralized training set | OASIS · CLARC Group 1 (standard) | MRR: 86.54 -> 87.48 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- CodeT5+ — CodeT: Code Generation with Generated Tests (ICLR 2023) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- CoRNStack: High-Quality Contrastive Data for Better Code Retrieval and Reranking (ICLR 2025 · similarity 0.86)
- CPRet: A Dataset, Benchmark, and Model for Retrieval in Competitive Programming (NeurIPS 2025 · similarity 0.86)
- CoRe: Benchmarking LLMs’ Code Reasoning Capabilities through Static Analysis Tasks (NeurIPS 2025 · similarity 0.85)
- LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code (ICLR 2025 · similarity 0.85)
- Mercury: A Code Efficiency Benchmark for Code Large Language Models (NeurIPS 2024 · similarity 0.85)
- SWE-fficiency: Can Language Models Optimize Real-World Repositories on Real Workloads? (ICML 2026 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
