Competition-level code generation with AlphaCode
Explore this paper's citation graph
Summary
AlphaCode is introduced, a system for code generation that achieved an average ranking in the top 54.3% in simulated evaluations on recent programming competitions on the Codeforces platform, marking the first time an artificial intelligence system has performed competitively in programming competitions.
- Type
- preprint
- Published
- 2022-02-08
- Cited by
- 2,455
- References
- 82
- Access
- Open access
- OpenAlex
- https://openalex.org/W4226287673
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:246527904
Keywords
Computer science, Code (set theory), Set (abstract data type), Programming paradigm, Artificial intelligence
References
- Learning from examples to improve code completion systems
- The rust language
- Toward automatic program synthesis
- Scratch
- Automating string processing in spreadsheets using input-output examples
- On the naturalness of software
- Code completion with statistical language models
- How Program History Can Improve Code Completion
- PPDB: The Paraphrase Database
- Program synthesis by sketching
- DeepCoder: Learning to Write Programs
- SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processing
- The adverse effects of code duplication in machine learning models of code
- Sequence to Sequence Learning with Neural Networks
- Hierarchical Neural Story Generation
- BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
- Latent Predictor Networks for Code Generation
- Achieving Verified Robustness to Symbol Substitutions via Interval Bound Propagation
- Program Synthesis
- Grandmaster level in StarCraft II using multi-agent reinforcement learning
Cited by
- Impossibility Results in AI: A Survey
- Less is More: Summary of Long Instructions is Better for Program Synthesis
- Static Prediction of Runtime Errors by Learning to Execute Programs with External Resource Descriptions
- Compositional Generalization and Decomposition in Neural Program Synthesis
- On the Effectiveness of Pretrained Models for API Learning
- Learning to Induce Causal Structure
- MP-CodeCheck: Evolving Logical Expression Code Anomaly Learning with Iterative Self-Supervision
- InCoder: A Generative Model for Code Infilling and Synthesis
- CodexDB: Generating Code for Processing SQL Queries using GPT-3 Codex
- ET-BERT: A Contextualized Datagram Representation with Pre-training Transformers for Encrypted Traffic Classification
- Natural Language to Code Translation with Execution
- Towards Teachable Reasoning Systems: Using a Dynamic Memory of User Feedback for Continual System Improvement
- An Initial Look at Self-Reprogramming Artificial Intelligence
- Memorizing Transformers
- A Conversational Paradigm for Program Synthesis
- From Solution Synthesis to Student Attempt Synthesis for Block-Based Visual Programming Tasks
- Improving automatically generated code from Codex via Automated Program Repair
- Towards Learning Universal Hyperparameter Optimizers with Transformers
- Formal Specifications from Natural Language
- Productivity assessment of neural code completion