Contrastive Code Representation Learning
Open accessAbstract
Recent work learns contextual representations of source code by reconstructing tokens from their context.For downstream semantic understanding tasks like code clone detection, these representations should ideally capture program functionality.However, we show that the popular reconstruction-based RoBERTa model is sensitive to source code edits, even when the edits preserve semantics.We propose Con-traCode: a contrastive pre-training task that learns code functionality, not form.Con-traCode pre-trains a neural network to identify functionally similar variants of a program among many non-equivalent distractors.We scalably generate these variants using an automated source-to-source compiler as a form of data augmentation.Contrastive pretraining outperforms RoBERTa on an adversarial code clone detection benchmark by 39% AUROC.Surprisingly, improved adversarial robustness translates to better accuracy over natural code; ContraCode improves summarization and TypeScript type inference accuracy by 2 to 13 percentage points over competitive baselines.All source is available at https://github.com/parasj/contracode.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Existing techniques for optimizing LLM behavior on code snippets are mainly driven by machine learning approaches and lack explanations of LLM preference or predictable results before actual generation.
“These optimization techniques are mainly driven by machine learning approaches, lacking explanations on preference of LLM or predictable results before the actual generation.”
From What Makes In-Context Examples Effective for Code Generation? (2026) · cites this paper as Jain et al. [34]
Cites
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 48,293 · see citing works
RoBERTa: A Robustly Optimized BERT Pretraining Approach
Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., Levy, O., Lewis, M., Zettlemoyer, L., & Stoyanov, V. (2019) · arXiv · doi:10.48550/arxiv.1907.11692 · Cited by 16,515 · see citing works
FaceNet: A unified embedding for face recognition and clustering
Schroff, F., Kalenichenko, D., & Philbin, J. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298682 · Cited by 15,267 · see citing works
Deep Contextualized Word Representations
Peters, M., Neumann, M., Iyyer, M., Gardner, M., Clark, C., Lee, K., & Zettlemoyer, L. (2018) · Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers) · doi:10.18653/v1/n18-1202 · Cited by 11,983 · see citing works
Representation Learning with Contrastive Predictive Coding
Oord, A. V. D., Li, Y., & Vinyals, O. (2018) · arXiv · doi:10.48550/arxiv.1807.03748 · Cited by 6,242 · see citing works
Unsupervised Feature Learning via Non-parametric Instance Discrimination
Wu, Z., Xiong, Y., Yu, S. X., & Lin, D. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00393 · Cited by 3,664 · see citing works
SIGNATURE VERIFICATION USING A “SIAMESE” TIME DELAY NEURAL NETWORK
BROMLEY, J., BENTZ, J. W., BOTTOU, L., GUYON, I., LECUN, Y., MOORE, C., SÄCKINGER, E., & SHAH, R. (1994) · Series in Machine Perception and Artificial Intelligence · doi:10.1142/9789812797926_0003 · Cited by 2,350 · see citing works
The adverse effects of code duplication in machine learning models of code
Allamanis, M. (2019) · Proceedings of the 2019 ACM SIGPLAN International Symposium on New Ideas, New Paradigms, and Reflections on Programming and Software · doi:10.1145/3359591.3359735 · Cited by 284 · see citing works
CodeBERT: A Pre-Trained Model for Programming and Natural Languages
Feng, Z., Guo, D., Tang, D., Duan, N., Feng, X., Gong, M., Shou, L., Qin, B., Liu, T., Jiang, D., et al. (2020) · arXiv · doi:10.48550/arxiv.2002.08155 · Cited by 243 · see citing works
Superoptimizer: a look at the smallest program
Massalin, H. (1987) · Proceedings of the second international conference on Architectual support for programming languages and operating systems · doi:10.1145/36206.36194 · Cited by 154 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
Evaluating Large Language Models Trained on Code
Chen, M., Tworek, J., Jun, H., Yuan, Q., Pinto, H. P. D. O., Kaplan, J., Edwards, H., Burda, Y., Joseph, N., Brockman, G., et al. (2021) · arXiv · doi:10.48550/arxiv.2107.03374 · Cited by 1,436 · see citing works
“In fact, ContraCode (Jain et al., 2020) leveraged the large space of functionally correct programs to train a contrastive code model, which improved model performance on tasks like type inference.”
BackgroundCode Llama: Open Foundation Models for Code
Rozière, B., Gehring, J., Gloeckle, F., Sootla, S., Gat, I., Tan, X. E., Adi, Y., Liu, J., Sauvestre, R., Remez, T., et al. (2023) · arXiv · doi:10.48550/arxiv.2308.12950 · Cited by 531 · see citing works
“…of programs, with specialized objectives for code deobfuscation (Lachaux et al., 2021), contrastive learning through semantic-preserving code transformations (Jain et al., 2021), leveraging Abstract Syntax Trees to learn tree-aware positional encodings (Shiv & Quirk, 2019; Peng et al., 2021).”
Retrieval-Based Prompt Selection for Code-Related Few-Shot Learning
Nashid, N., Sintaha, M., & Mesbah, A. (2023) · 2023 IEEE/ACM 45th International Conference on Software Engineering (ICSE) · doi:10.1109/icse48619.2023.00205 · Cited by 178 · see citing works
ReACC: A Retrieval-Augmented Code Completion Framework
Lu, S., Duan, N., Han, H., Guo, D., Hwang, S.-W., & Svyatkovskiy, A. (2022) · arXiv · doi:10.48550/arxiv.2203.07722 · Cited by 96 · see citing works
“There exists several attempts to apply such transformation to code (Jain et al., 2020; Rabin et al., 2021; Bui et al., 2021).”
BackgroundWhat Makes In-Context Examples Effective for Code Generation?
Li, D., Chen, S., Cao, J., & Cheung, S.-C. (2026) · Proceedings of the ACM on Software Engineering · doi:10.1145/3832275
- Authors
- Jain, P., Jain, A., Zhang, T., Abbeel, P., Gonzalez, J., & Stoica, I.
- Affiliations
- University of California, Berkeley (US)
- Venue
- Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing
- Published
- 2021
- Pages
- 5954-5971
- OpenAlex
- W3040955294
- Cited by
- 129
- References
- 10
- Citations by year
- 2026: 6 · 2025: 24 · 2024: 25 · 2023: 34 · 2022: 26 · 2021: 12 · 2020: 2
- FWCI
- 20.11
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Conference on Empirical Methods in Natural Language Processing (EMNLP) (publishedVersion, cc-by)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- National Science Foundation (1730628, 1752814)
- Nvidia
- Amazon Web Services
- VMware
- Futurewei Technologies
- Ant Financial Services Group
- Directorate for Computer and Information Science and Engineering
- Office of Naval Research
- Topics
- Physical Sciences › Computer Science › Information Systems › Software Engineering Research
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Physical Sciences › Computer Science › Computer Science Applications › Teaching and Learning Programming
- Keywords
- contrastive learning, code representation learning, code clone detection, adversarial robustness, source-to-source compiler, data augmentation, code summarization
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
