Contrastive Code Representation Learning

Jain et al., Contrastive Code Representation Learning, 2020. For many machine-aided programming tasks, programs with the same functionality should have the same underlying representation. ContraCode learns such representations with contrastive learning: the network is trained to find equivalent programs among many distractors.

Contrastive Code Representation Learning Contrastive Code Representation Learning Paras Jain* UC Berkeley Ajay Jain* UC Berkeley Tianjun Zhang UC Berkeley Pieter Abbeel UC Berkeley Joseph Gonzalez UC Berkeley Ion Stoica UC Berkeley Paper (arXiv) GitHub For many machine-aided programming tasks, programs with the same functionality should have the same underlying representation . ContraCode learns such representations with contrastive learning: the network is trained to find equivalent programs among many distractors. Summary Developer tools increasingly use machine learning to understand and mo

saved by

• Martin Camacho

related reading

• [2203.07814] Competition-Level Code Generation with AlphaCodearxiv.org
• Contrastive Learning as Goal-Conditioned Reinforcement Learningarxiv.org
• Self-CTRL: Self-Consistency Training with Reinforcement Learningarxiv.org
• [2506.06266] Cartridges: Lightweight and general-purpose long context representations via self-studyarxiv.org
• Discovering Concept-Editing Algorithms With LLM Agentsdmodel.ai
• [2510.05056] Modeling Student Learning with 3.8 Million Program Tracesarxiv.org
• [2510.27062] Consistency Training Helps Stop Sycophancy and Jailbreaksarxiv.org
• Inverse Constitutional AI: Compressing Preferences into Principlesarxiv.org
• Towards Automated Circuit Discovery for Mechanistic Interpretabilityarxiv.org
• pdfopenreview.net
• Contrastive Representation Learning | Lil'Loglilianweng.github.io
• Sparse Autoencoders Find Highly Interpretable Features in Language Modelsarxiv.org